arX

iv:1

707.

0930

0v2

[cs

.NI]

16

Jan

2018

1

From 4G to 5G: Self-organized Network

Management meets Machine Learning

Jessica Moysen1 and Lorenza Giupponi2

1Department of Signal and Theory Communications, Universitat Politecnica de Catalunya-UPC

Email: jessica.moysen@tsc.upc.edu

2Communications Network Division, Centre Tecnologic de Telecomunicacions de Catalunya-CTTC

Email: lorenza.giupponi@cttc.es

Abstract—In this paper, we provide an analysis of self-organized network management, with an end-to-end perspectiveof the network. Self-organization as applied to cellular networksis usually referred to Self-organizing Networks (SONs), and itis a key driver for improving Operations, Administration, andMaintenance (OAM) activities. SON aims at reducing the cost ofinstallation and management of 4G and future 5G networks, bysimplifying operational tasks through the capability to configure,optimize and heal itself. To satisfy 5G network managementrequirements, this autonomous management vision has to beextended to the end to end network. In literature and also in someinstances of products available in the market, Machine Learning(ML) has been identified as the key tool to implement autonomousadaptability and take advantage of experience when makingdecisions. In this paper, we survey how network managementcan significantly benefit from ML solutions. We review andprovide the basic concepts and taxonomy for SON, networkmanagement and ML. We analyse the available state of the artin the literature, standardization, and in the market. We payspecial attention to 3rd Generation Partnership Project (3GPP)evolution in the area of network management and to the datathat can be extracted from 3GPP networks, in order to gainknowledge and experience in how the network is working, andimprove network performance in a proactive way. Finally, we gothrough the main challenges associated with this line of research,in both 4G and in what 5G is getting designed, while identifyingnew directions for research.

Index Terms—Network Management, Machine Learning, Self-Organizing Networks, Mobile Networks, Big Data

I. INTRODUCTION

Traditionally, and up to 4G, the evolution from one gen-

eration of mobile networks to another, has been driven by

hardware technology advancements. The revolution of 5G

is different, and novel advancements of software technology

will be critical, especially in the way the network will be

managed. With the advent of these software advancements,

and unprecedented levels of computational capacity, the vision

of autonomous network management can be put into practice

taking advantage of also other cross-disciplinary knowledge

advancements in the area of Machine Learning. This vision

aligns with the concepts of self-awareness, self-configuration,

self-optimization, and self-healing, which have already been

The research leading to these results has received funding from the SpanishMinistry of Economy and Competitiveness under grant TEC2014-60491-R(Project 5GNORM). This work also was supported by the Spanish NationalScience Council and ERFD funds under Project TEC2014-60258-C2-2-R.

defined in the area of network management. We give an

special emphasis to the access segment of 4G cellular Long

Term Evolution (LTE) network through the concept of Self-

organizing Network (SON). SON is a common term used to

refer to mobile network automation and minimization of hu-

man intervention in the cellular/wireless network management.

This concept has been introduced by 3GPP in Release 8 and

it has been expanding across subsequent releases. 3GPP work

is inspired by a set of requirements defined by the opera-

tors’ Alliance Next Generation Mobile Networks (NGMN).

The main objective of SON can be roughly classified into

three main points: 1) to bring intelligence and autonomous

adaptability into cellular networks, 2) to reduce capital and

operation expenditures (CAPEX/OPEX), 3) to enhance net-

work performances in terms of network capacity, coverage,

offered service/experience, etc. SON is considered today as a

driving technology that aims at improving spectral efficiency,

simplifying management, and reducing the operation costs

of the next generation Radio Access Networks (RANs). The

overall complex SON problem has been decomposed in a set

of useful use cases, which have been identified by 3GPP,

NGMN, 5G Infrastructure Public Private Partnership (5GPPP)

and different EU projects [1]–[6]. The academic literature has

dedicated significant effort to SON algorithms in the context

of the above mentioned use cases, providing smart solutions to

optimize network operator performance, expenses and users’

experience. Many of these works are already reviewed here

[7]. The market also offers already complete sets of SON so-

lutions, (e.g. [8]–[11], among others) many products have been

advertised and presented in Mobile World Congress (MWC)

2016 and 2017 [12], [13]. For example, AirHop’s eSON

from Jio & AirHop communications [11], which employs a

multi-vendor, multi-technology, real-time SON solution based

on scalable and virtualized software platform has recently

been awarded for the 2016 Small Cell Forum Heterogenous

Network (Het-Net) management software and service award

[14].

However, to the best of our knowledge, the SON solutions

available in the market are 1) mainly based on heuristics,

2) the automated information processing is usually limited to

low complexity solutions like triggering, 3) many operations

are still done manually (e.g. network faults are usually fixed

directly by engineers), 4) SON solutions do not really capi-

2

talize on the huge amount of information that is available in

mobile networks to build next generation network management

solutions, and 5) several open challenges are still unsolved,

like the problem of coordination of SON functions [15], [16],

or the proper solution of the trade-off between centralized and

distributed SON implementations [17]. In addition, this self-

organized network management vision should be extended also

beyond the RAN segment and should include all the segments

of the network, from the access to the core, while fulfilling

the requirements of different kind of vertical service instances.

To achieve this vision, the networking world is exploring

new directions. Network Functions Virtualisation (NFV) is

expected to bring the economy of scale of the Information

Technology industry to the Telecom industry. When combining

NFV with Software Defined Network (SDN) principles, the

benefits of programmability and flexibility are brought to the

fore.

Another aspect that should be considered is that, as we

observed in [18] a huge amount of data is currently already

generated in 4G networks during normal operations by control

and management functions, and more is expected to come

in 5G networks due to the densification process [19], het-

erogeneity in layers and technologies, the additional control

and management complexity in NFV and SDN architectures,

the advent of Machine to Machine (M2M) and Internet of

Things (IoT) paradigms, the increasing variety of application

and services, each with distinct traffic patterns and Quality of

Service (QoS)/Quality of Experience (QoE) requirements, etc.

5G network management is expected to provide a whole new

set of challenges due to 1) the need to manage future network

complexity, due to ultra dense deployments, heterogeneous

nodes, networks, applications, RANs coexisting in the same

setting, 2) the need to manage very dynamic networks, where

operators may do not have any control in the deployment

of some nodes (e.g. femto-cells), energy saving policies are

in place generating a fluctuating number of nodes, active

antennas are a reality, etc. 3) the need to support 1000x

traffic, and 10x users, and improve energy efficiency, 4) the

need to improve the experience of the users by enabling

Gbps speeds, and highly reduced latency, 5) the need to

manage new virtualized architectures, 6) the need to handle

heterogeneous spectrum access privileges through novel LTE-

Unlicensed (LTE-U), Licensed Assisted Access (LAA), MuL-

TEFire paradigms and the availability of both traditional sub

- 6 GHz bands, and above 6 GHz mmWave bands. In this

challenging context, we believe that the use of SON and of

smart network management policies is crucial and inevitable

for operators running multi-RAT, multi-vendor, multi-layer

networks, where an overwhelming number of parameters need

to be configured and optimized. The high level objective is

to make the networks 1) more self-aware, by exploiting the

information already available in the network to gain experi-

ence in the network management, 2) more self-adaptive, by

exploiting intelligent control decisions procedures which allow

to automatize the decision processes based on the experience.

We believe that Machine Learning (ML) can be effectively

used to allow the network to learn from experience, while im-

proving performance. In particular, big data analytics, through

the analysis of data already generated by the network, can

pursue the self-awareness by driving the network management

from reactive to predictive. Big data analytics are currently

receiving big attention in research and in the market, due to

their capability of providing insightful information from the

analysis of data already available to operators.

In this paper, we will not focus on these uses of data

analytics and ML, but we will only focus on the applica-

tions to 4/5G network management. Differently from other

surveys on SON proposed before [7], [20] or from other

surveys related with 5G network management [21], we focus

here on the study and analysis of the available literature on

SON and network management considering ML as the tool

to implement automation and self-organization, from a 5G

perspective. We review and provide the basic concepts and

taxonomy of traditional SON and 5G network management

in Section II. We pay special attention to the evolution of

3GPP in the area, following its nomenclature, and referring to

the specific use cases defined by the standard in this matters.

We then provide, in Section III, guidelines to select the most

appropriate ML algorithm and approach, based on the network

management issue to address. In Section IV, we review the

main sources of information relevant for a knowledge based

network management, data that is actually already generated

by the network, and we survey the literature on ML-based

network management. We then highlight challenges for future

works in Section V. Finally, Section VI concludes the survey.

II. SELF ORGANIZING NETWORK (SON) AND NETWORK

MANAGEMENT

SON is a key driver to maximize total performance in

cellular networks. The main idea is to bring into them in-

telligence and autonomous adaptability by diminishing human

involvement, while enhancing network performance, in terms

of network capacity, coverage and service quality. It aims at re-

ducing the cost of installation and management by simplifying

operational tasks through the capability to configure, optimize

and heal itself.

The main motivation behind the increasing interest in the

introduction of SON from operators, standardization bodies

and projects is twofold. On the one hand, from the market

perspective, the ever increasing demand for a diversity of

offered services, and the need to reduce the time to market

of innovative services, further add to the pressure to remain

competitive by effectuating cost reductions. On the other hand,

from the technical perspective, the complexity and large scale

of future radio access technologies imposes significant oper-

ational challenges due to the multitude of tunable parameters

and the intricate dependencies among them. In addition, the

advent of heterogeneous networks is expected to tremendously

increase the number of nodes in this new ecosystem, so classic

manual and field trial design approaches are just impractical.

Similarly, manual optimization processes or fault diagnosis

and cure, performed by experts are no longer efficient and need

to be automatized, as this causes time intensive experiments

with limited operational scope, or delayed, manual and poor

handling of cell/sites failures. Key operational tasks, such

3

Self-Configuration

(ANR, automatic PCI assignment)

Self-Healing

(recovery of NE, SW, COD, COC)

Self-Optimization

(MLB, MRO, ICIC, RACH opt., CCO, ES)

Planning

Deployment

Maintenance/

Optimization

Fig. 1: Self-organizing networks

as radio network planning and optimization are largely sep-

arated nowadays and this causes intrinsic shortcomings like

the abstraction of access technologies for network planning

purposes, or the consideration of performance indicators that

are of limited relevance to the end user’s service percep-

tion. These problems have been approached through SON

by European projects such as SOCRATES [2], and Gandalf

[1]. Also FP7 and 5GPPP EU projects have been dealing

with SON. In particular FP7 SEMAFOUR [3], which devel-

ops a unified self-management to operate complex HetNets.

Among 5GPPP projects, we highlight SESAME [4], which

proposes the Cloud-Enabled Small Cell (CESC) concept,

i.e., a new multi-operator enabled Small Cell by deploying

Network Functions Virtualisation (NFV), supporting powerful

self-management inside the access network infrastructure. In

terms of self-organized network management in SDN and

NFV, the article project aims at enabling the use of ML to

achieve real time autonomous 5G network management [5]. In

particular, the project explores a smart integration of state-of-

the-art technologies in SDN, NFV), SON, Cloud computing,

Artificial intelligence. The COGNET project [6] has similar

objectives and aims at developing several operators’ use cases

by applying ML algorithms.

SON has been introduced by 3GPP as a key component

of LTE network starting from the first release of this tech-

nology in Release 8, and expanding to subsequent releases. In

SOCRATES [2] and in 3GPP [22], meaningful SON use cases

have been defined, which can be classified according to the

phases of the life cycle of a cellular systems (planning, deploy-

ment, maintenance and optimization) into: self-configuration,

self-healing and self-optimization, as depicted in Figure 1. In

this section, first we give an overview of the evolution of

SON in 3GPP. We go through self-configuration, optimization

and healing functionalities, introducing the use cases that have

been defined for each one of them. We discuss about the self-

coordination problem, to handle the potential conflicts that

may exist between the parallel execution of multiple SON

functions. We present the Minimization of Drive Tests (MDT)

functionality. Finally, we focus on and end-to-end vision by

extending SON principles to the core, and we discuss the role

of virtualized and software defined networks in the context of

5G Network Management. Notice that here we do not focus

on the academic literature, as it has already been reviewed in

other interesting works [7]. We focus on the taxonomy defined

by 3GPP, on the related roadmap, and we pay attention to the

market penetration.

A. SON evolution in 3GPP

3GPP Release 8 started defining LTE and already sets the

basis for concepts and requirements, and for SON functionali-

ties regarding self-configuration, initial equipment installation

and integration. The ANR functionality is introduced here to

reduce manual work when configuring the neighbouring list

in newly deployed eNBs. Concepts of self-optimization are

defined in the context of Release 9. It includes optimisation

of coverage, capacity, handover and interference. The func-

tions which are introduced (and that will be detailed in the

following sections) are Mobility Load Balancing (MLB), Mo-

bility Robustness/Handover Optimisation (MRO), Inter-Cell

Interference Coordination (ICIC) and Random Access Channel

(RACH) optimization. Release 10 focuses on enhancements

to already defined SON functions to enhance interoperability

between small cells and macro-cells and includes NGMNs

recommendations, i.e., new functionalities such as Coverage

and Capacity Optimization (CCO), enhanced ICIC, and it

defines all the concepts related to self-healing, so Cell Out-

age Detection (COD) and Cell Outage Compensation (COC)

functions. Finally, concepts of MDT and Energy Saving (ES)

are also introduced and then enhanced in Release 11. Release

11 SON functions are related to the automated management

of heterogeneous networks. It includes mobility robustness

optimization enhancements and inter-radio access technology

Handover (HO) optimization. Release 12 introduces optimiza-

tion and enhancements for small cells including deployments

in dense areas. In Release 13 novel concepts of unlicensed

LTE have been introduced. Besides that, Release 13 studied

the enhancements of OAM, with respect to centralized and

distributed architecture. In particular, focuses on distributed

MLB, as well as on enhanced NM or centralized CCO. Finally,

Release 14 focuses on meeting the 5G requirements in terms of

latency reduction, use of unlicensed spectrum in a fair manner,

support for carrier aggregation, energy efficiency at OAM

level, SON for active antennas, etc. Table I summarizes the

evolution of SON in 3GPP. Other documents of interest also

include the protocol neutral SON policy Network Resource

Model (NRM) Integration Reference Point (IRP), with the

Information Service (IS) [31], [41] and Solution Sets (SS) [42],

[43].

B. Self Configuration

Self-configuration is the process of bringing a new network

element into service with minimal human operator intervention

[24]. This covers the cellular system life cycle phase related

to planning and deployment. Self-configuring algorithms take

care of all configuration aspects of the Enhanced Node Base

station (eNB). When the eNB is powered on, it detects the

transport link and establishes a connection with the core

network elements. After this, the eNB is ready to establish

4

TABLE I: Evolution of SON in 3GPP

Release WI Feature TS or TR

Rel.8 SA5-SON concepts and requirements SON concepts and requirements [23]

Rel.8 SA5-Self establishment of eNBs Self configuration [24]–[29]

Rel.8 SA5-SON Automatic Neighbour Relation (ANR) list management ANR, PCI [30]–[33]

Rel.9 SA5: Study of SON related OAM Interfaces for HeNBs SON related OAM Interfaces for HeNBs [34]

Rel.9 SA5: Study of self-healing of SON Self-healing management [35]

Rel.9 SA5:SON OAM aspects: Automatic radio network configuration datapreparation

Automatic radio network configuration data prepara-tion

[24]–[26]

Rel.9 SA5:SON OAM aspects self-organization management Self-optimization (MRO, MLB, ICIC) [36]

Rel.9 RAN3: Self-organizing networks CCO, MRO, MLB, RACH opt. [37]–[40]

Rel.10 SA5: SON self-optimization management continuation Self-coordination, self-optimization (MRO, MLB,ICIC, RACH opt.)

[31], [36], [41]–[43]

Rel.10 SA5: Self-healing management CCO, COC [44]

Rel.10 SA5: OAM aspects of ES in radio networks ES [31]–[33], [36], [45]–[47]

Rel.10 RAN2-3: LTE SON enhancements CCO, ES, MLB, MRO enhancements [38]–[40], [48]

Rel.11 SA5: ULTRAN SON management SON management [23], [41], [42], [49]–[52]

Rel.11 SA5: LTE SON coordination management SON coordination [53] [23], [31], [36], [41], [42], [51]

Rel.11 SA5: Inter RAT ES management OAM aspects of ES management [41], [42], [47], [49], [52], [54]

Rel.11 RAN3: Further SON enhancements MRO, MDT enhancements [37]–[40], [48], [55]–[58]

Rel.12 SA5: Enhanced NM centralized CCO Enhanced NM centralized CCO [36], [59]–[64]

Rel.12 SA5: Multi-vendor plug and play eNB connection to the network Multi-vendor plug and play eNB connection to thenetwork

[24], [65], [66]

Rel.12 SA5: Enhancements on OAM aspects of distributed MLB OAM aspects of distributed MLB [67]

Rel.12 SA5: Energy efficiency related performance measurements Energy efficiency related performance measurements [36]

Rel.12 SA5: Het-Nets management/OAM aspects of network sharing Het-Nets/network sharing [68], [69]

Rel.12 RAN2-3: Next generation SON for ULTRAN/EUTRAN SON per UE type, active antennas, small cells [70]

Rel.12 RAN2-3: ES enhancements for EUTRAN ES [71]

Rel.13 RAN2-3: Enhanced Network Management centralized CCO CCO [61]

Rel.13 SA5: Study on Enhancements of OAM aspects of Distributed MobilityLoad Balancing SON function

MLB [72]

Rel.14 RAN: OAM (SON for Active Antenna Systems (AAS)-based deploy-ments)

Energy efficiency [73], [74]

Fig. 2: SON implementations.

OAM, S1 and X2 links and finally sets itself in operational

mode. After the eNB is configured, it performs a self-test to

deliver a status report to the network management node. Since

Release 8 ANR and Automated Configuration of Physical Cell

Identity (PCI) use cases have been considered [75], [76]. The

ANR function resides in the eNB and manages the conceptual

Neighbour Relation Table (NRT). Located within ANR, the

Neighbour Detection Function finds new neighbours and adds

them to the NRT. ANR also contains the Neighbour Removal

Function which removes outdated NRs. The Neighbour De-

tection Function and the Neighbour Removal Function are

implementation specific [77]. The PCI is a physical layer

signature to distinguish signals from different eNBs. It is

based on synchronization signals. The total number of PCIs

is LTE is 504, so that reuse is inevitable, especially in

dense deployments. The Automatic PCI assignment aims at

an automatic conflict and confusion free identification of cells

[78]. Recommended practices for both use cases can be found

in [79].

C. Self Optimization

Self-optimization embraces all the set of mechanisms which

optimize the network parameters during operation, based on

measurements received from the network. In the following we

provide a brief overview of the main self-optimization function

that have been introduced across the different recent releases

[78]. From Release 9, we highlight work on:

1) MLB. The MLB is the SON function in charge of

managing cells’ congestion through load transfer to

other cells. The main objective is to improve the end-

user experience and achieve higher system capacity

by distributing user traffic across the system radio re-

sources. The implementation of this function is gener-

ally distributed and supported by the load estimation

and resource status exchange procedure. The messages

containing useful information for this SON function

5

eNB2

O&M

Itf-S Itf-S

UE Measurements

RSRP (Ref. Symbol Received Power)

RSRQ (Ref. Symbol Received quality)

CQI (Channel Qaulity Indicator)

Action request 1: Adjust antenna parameters by

RET for coverage improvement

Action request 2: Increase TXP for coverage

improvement

Action request 3: Decrease TXP

for interference reduction

1. Resource conflict between MRO and MLB

2. COC and ICIC resource conflict

3. CCO and ICIC resource conflict

eNB1

ICIC ICIC

X2

RLF (Radio Link Monitoring)

HO (Handover) report

Mobility Change

Action request 6: Postpone HO to

avoid ping pong effect

Parameters to change: Longer

offset Δ1, larger TTT (Time to

Trigger)/FC (Filter Coefficient)

Action request 7: Postopone

HO to favour LB

Parameters to change: Longer

offset Δ2, larger /FC

tImpact time

MLBResource status request and

response

Find cell/UE candidate for LB

(Load Balancing)

Execution HO procedure

Mobility change

MRO MRO coorection

MRO root cause

evaluation

MRO MRO coorection

MRO root cause

evaluation

COC

COC

Action request 4: Adjust antenna

parameters by RET to solve outage

problems

Action request 5: Increase TXP to

solve outage problemsCCO

COC

MLB

Fig. 3: High-level example of how the iterations of multiple SON functions may interfere.

(resource status request, response, failure and update)

are transmitted over the X2 interface [40]. MLB can be

implemented by tuning the Cell Individual Offset (CIO)

parameter. The CIO contains the offsets of the serving

and the neighbour cells that all UEs in this cell must

apply in order to satisfy the A3 handover condition [80].

2) MRO. The MRO is a SON function designed to guaran-

tee proper mobility, i.e. proper handover in connected

mode and cell re-selection in idle mode. Among the

specific goals of this function we have the minimization

of call drops, the reduction of Radio Link Failures

(RLFs), the minimization of unnecessary hand-overs,

ping-pongs, due to poor handover parameters settings,

the minimization of idle problems. Its implementation is

commonly distributed. The messages containing useful

information are: the S1AP handover request or X2AP

handover request, the handover report, the RLF indi-

cation/report. Release 11 focused on different improve-

ments of the handover optimization [81]. MRO operates

over connected mode and idle mode parameters. In

connected mode, it tunes meaningful handover trigger

parameters, such as the event A3 offset (when referring

to intra-RAT, intra-carrier hand-overs), the Time to Trig-

ger (TTT), or the Layer 1 and Layer 3 filter coefficients.

In idle mode, it tunes the offset values, such as the Qoff-

set for the intra-RAT, intra-carrier case.

3) Inter-Cell Interference Coordination (ICIC). ICIC aims

to minimize interference among cells using the same

spectrum. It involves the coordination of physical re-

sources between neighbouring cells to reduce interfer-

ence from one cell to another. ICIC can be done in

both uplink and downlink for the data channels Physical

Downlink Shared Channel (PDSCH), and Physical Up-

link Shared Channel (PUSCH), or uplink control channel

Physical Downlink Control Channel (PDCCH). ICIC can

be static, semi-static or dynamic. Dynamic ICIC relies

on frequent adjustments of parameters, supported by

signalling among cells over X2 interface. To support

proactive coordination among cells the High Interference

Indicator (HII) and the Relative Narrowband Transmit

Power (RNTP) indicators have been defined, while to

support reactive coordination, the Overload Indicator

(OI) has been introduced [40].

4) RACH. RACH optimization aims at optimizing the ran-

dom access channels in the cells based on UE feedback

and knowledge of its neighbouring eNBs RACH config-

uration. RACH optimization can be done by adjusting

the Power control (Pc) parameter or change the preamble

format to reach the set target access delay [82].

In Release 10, 3GPP defined new use cases.

1) Coverage and Capacity Optimization (CCO) is a SON

function that aims to design self-optimizing algorithms

that achieve optimal trade-offs between coverage and

capacity. Different mechanisms can be considered to

dynamically improve coverage and capacity, such as

ICIC, scheduling, and the combination of such mecha-

nisms. The targets that can be optimized may be vendor

dependent and include coverage, cell throughout, edge

cell throughput, or a weighted combination of the above.

2) ES aims at providing the quality of experience to end

users with minimal impact on the environment. The

objective is to optimize the energy consumption, by

designing Network Elements (NEs) with lower power

consumption and temporarily shutting down unused

capacity or nodes when not needed [45]. In particu-

lar, many works in literature have been focusing on

switching ON/OFF eNBs or small cells, in an efficient

way, in order to guarantee a target level of Quality

6

of service/experience, while minimizing the dissipated

energy.

Release 11 provides enhancements to MLB optimization, HO

optimization, CCO, and ES. Release 12 has focused on a study

on enhancements of OAM aspects for distributed MLB [72].

D. Self-healing

Self-healing [44] focuses on the maintenance phase of

a cellular network. Wireless cellular systems are prone to

faults and failures, and the most critical domain for fault

management is the RAN. Every eNB is responsible for serving

an area, with little or none redundancy. If a NE is not able to

fulfill its responsibilities, it results in a period of degradation of

performances, during which users are not receiving a proper

service. This results in severe revenue loss for the operator.

Self-healing was initially studied in Release 9 [35], but it

is in Release 10, when the main work has been carried out

and features for detection, and adjustment of parameters have

been specified [83]. These specifications have been further

updated in Release 11 [44]. The main defined use cases are

the following.

1) Self-recovery of NE Software. If the NE software failed

due to load earlier software version and/or configu-

ration, the most important thing is to ensure that the

NE runs normally by removing the fault software, and

restoring the configuration.

2) Self Healing of board Faults. This use case aims to solve

hardware failures in the NE [84].

3) Cell Outage Management. This use case is split in two

main functions: 1) Cell Outage Detection. The main

objective here is to detect a cell outage through the

monitor performance indicators, which are compared

against thresholds and profiles, and 2) Cell Outage

Compensation. This use case aims at alleviating the

outage caused by the loss of a cell from service [44].

It refers to the automatic mitigation of the degradation

effect of the outage by appropriately adjusting suitable

radio parameters, such as the pilot power and the antenna

parameters of the surrounding cells.

E. Self Coordination

SON functionalities are often designed as stand-alone func-

tionalities, by means of control loops. When they are executed

concurrently in the same or different network elements, the

impact of their interactions is not easy to be predicted, and

unwanted effects may even occur among instances of the same

SON function, when implemented in neighbouring cells. The

risk of unacceptable oscillations of configuration parameters

or undesirable performance results increase with the number

of SON functions.

3GPP has proposed different architectures for SON imple-

mentation, ranging from centralized C-SON to distributed D-

SON. The choice of the architecture has a strong impact on

the efficiency of the self-coordination framework. If C-SON

is used, SON functions are implemented in the Operation and

Maintenance Center (OMC) or in the Network Management

Systems (NMS), as part of the Operation and Support System

(OSS). This implementation benefits from global information

about metrics and Key Performance Indicator (KPI)s, as well

as computational capacity to run powerful optimization algo-

rithms involving multiple variables or cells. However, it suffers

from long time scales. In order to avoid oscillations of decision

parameters, 3GPP requires [53] that each SON function asks

for permission before changing any configuration parameter.

This means that a request must be sent from the SON function

to the SON coordinator and a response has to be returned.

In Centralized SON (C-SON) all these requests must pass

through the Interface-N, which is not suitable for real-time

communication, so that there is no possibility to give priority

to SON coordination messages over other OAM messages.

If in turn, distributed coordination is used, the interaction

between the SON function and the local SON coordinator

will be over internal vendor-specific interfaces, with much

lower latency characteristics. This makes the Distributed SON

(D-SON) architecture much more flexible and adequate for

small cell networks, which experience very transitory traffic

loads, thus requiring high reactivity to propagation and traffic

conditions.

An example of this can be observed in Figure 3, where

we provide an analysis of how the iterations among several

SON functions implemented in centralized and distributed

manner can generate conflicts in the network. In particular,

this figure focuses on the SON output parameter conflict, i.e.,

when two or more SON functions aim at optimizing the same

output parameter with different actions request, and where at

least three possible conflicts can arise: 1) the resource conflict

between MRO and MLB; 2) the one among CCO and ICIC,

and/or 3) the one among COC and ICIC use cases. We can

identify output parameters, which are affected by two opposite

decisions of two different functions, trying to achieve their

own targets. As a result, to define and implement a self-

coordination framework is considered a necessity [2], [85],

[86].

Market implementations of C-SON are offered by vendors

like Celcite (acquired by AMDOCS), Ingenia Telecom and

Intucell (acquired by Cisco), while D-SON solutions have

traditionally been more challenging to implement and vendor

specific, not allowing for easy interaction of products from

different vendors, so that a supervisory layer is commonly still

needed to coordinate the different instances of D-SON across

a much broader scope and scale. Only recently, vendors like

Qualcomm or Airhop have started proposing D-SON as a SON

mainstream, as small cells and Het-Net require the millisecond

response times of D-SON.

F. Minimization of Drive Tests

MDT enables operators to collect User Equipments (UEs)

measurements together with location information, if avail-

able, with the purpose of optimizing network management

while reducing operational effects and maintenance costs.

This feature has been studied by 3GPP since Release 9 [87],

among the targets there are the standardization of solutions

for coverage optimization, mobility, capacity optimization,

7

parametrization of common channels, and QoS verification

[84]. Since operators are also interested in estimating QoS

performance, in Release 11, MDT functionality has been

enhanced through QoS performance to properly dimension

and plan the network by collecting measurements indicating

throughput and connectivity issues [88]. These MDT functions

have been further elaborated in Release 11, while Release 12

has included specific enhancements in terms of correlation of

information, which can be found in the study on enhanced

network management centralized CCO. These improvements

and extensions of SON enhancements introduced until Release

13 can be found in [23].

G. Core networks

The core network operations can be managed through

self-organizing functionalities. The benefits also in this case

come from the reduced human intervention and from reduced

operational costs. self-organization in the core network allows

to self-adapt traffic loads and prevent bottlenecks. In addition,

self-organization for Core enables the core network to handle

signalling more efficiently. In this regard, Nokia [89] already

automates core networks operations based on SON technology.

The objective is to automatically and rapidly allocate core net-

work resources to meet unpredictable behaviours and demands

in terms of broadband. Notice that SON use cases for core

networks are not limited to LTE networks, but many of them

can be taken into account also for other kinds of networks,

like 2/3G.

H. Virtualized and Software defined networks

The wireless industry is currently working towards being

prepared for a 1000x data traffic growth. It is unlikely,

though, that users will want to pay more for the service

than they are paying today, which set a serious challenge

for both mobile operators and vendors, i.e. how to improve

the infrastructure 1000 times, without increasing the CAPEX

and OPEX. Besides SON, another trend in this direction,

initiated by an ETSI industrial study group in 2012, is the

NFV, which allows to exploit the economies of scale of the IT

industry, by moving traditional network functions away from

specialized hardware to general purpose computation, storage

and memory pools, distributed throughout the network and in

data centers. NFV virtualizes the functional elements of the

network, instantiating the corresponding functions as programs

that run on commercial off-the shelf, and less expensive

hardware. This concept, combined with a SDN architecture, is

introduced to make mobile network deployments more cost-

effective [21], [90].

The main idea behind these novel architectures is to pro-

vide a framework capable of assisting network operators to

solve management problems, such as, cyber attacks, network

failures, optimization to improve network performance, and

QoE of the users, among others. In this context, SON can be

useful to achieve real time autonomous network management.

In this novel softwerized visions, we can benefit of all the

opportunities offered by centralized, distributed and local

implementations proposed for SON at RAN level, to extend

this view beyond the radio access border, by proposing a

SON over NFV architecture, where SON functions, aimed

at tackling the main radio access and backhauling challenges

of extremely dense deployments, are virtualized and run over

generic purpose hardware. The NFV infrastructure is to be

managed by an orchestrator entity, as proposed in ETSI

architecture. Out of all the NFV architecture entities, this

is the brain with the broadest view of the vertical service

characteristics and the resource availability in the network.

Therefore, it coordinates the allocation of functions across the

different segments of the dense, heterogeneous network. At

the methodological level, the orchestrator can take advantage

of the huge amount of information travelling through the

network, in terms of measurements, signaling information,

QoS and QoE indicators, etc., by means of machine learning

based approaches.

In the market there already exist start-ups which advertise

the concept of C-SON in the cloud. SON over NFV eliminates

software and hardware dependencies, besides system scaling

limitations, and offers reduction of costs through automatic

processes. Cellwize [12] is one of them. They are promising a

technology with deployment in the cloud, capable of working

seamlessly across different vendors, spectrum and technolo-

gies. This research line is extremely novel and not much work

can be found so far. However, we highlight the work that is

under development in the context of the article and COGNET

projects [5], [6].

III. HOW TO ADDRESS SON AND NM THROUGH ML

In this section we classify at high level the different net-

work management classes of problems that one may need to

deal with when aiming at managing the network in a self-

organized manner. For each class of problem, we identify the

machine learning tools that can be used. The objective of

ML is to improve performance of a particular sets of tasks

by creating a model that helps find patterns through learning

algorithms. ML taxonomy is traditionally organized onto: 1.

Supervised Learning (SL), 2. Unsupervised Learning (UL),

3. Reinforcement Learning (RL). Recently, new trends in the

area of ML are taking momentum, thanks to the progress of

software engineering, computational capabilities and memory

availability. Deep learning has been proven feasible and ex-

tremely effective in different applications, like language, video,

speech recognition, object and audio detection, among others.

The most exemplary one is the win of AlphaGo, beating

the world champion at the Chinese board game Go. The

victory of AlphaGo was due to the implementation of a deep

reinforcement learning algorithm capable of self-learning.

Keeping in mind the SON and NM functions introduced in

the previous section, the classes of problems that need to be

addressed when managing the network autonomously are:

• Variable estimation or classification: The tasks belonging

to this class of problem aim at e.g. estimating the QoS

or the QoE of the network, at predicting performances or

behaviours of the network, by learning from the analysis

of data obtained from past behaviours of the network. NM

and SON functions where these tasks are useful are QoS

8

estimation and other MDT use cases, the prediction of

behaviours to optimize network parameters, etc. Solutions

to these problems can be translated into finding the

relationship between one variable and some others, or

Identifying which class of a set of pre-defined classes

the data belongs to. Solutions are then to be found in

the SL literature, with both regression and classification

tasks.

• Diagnosis of network faults or misbehaviours: The tasks

belonging to this class of problems aim at detecting issues

ongoing in the network, which may be associated to faults

and anomalous setting of network parameters. This kind

of problems relates to self-healing issues and solutions

can be found in UL literature, and in particular in the

anomaly detection solution.

• Dimensionality reduction: The network generates contin-

uously a huge amount of data. For an appropriate process-

ing and to extract useful information, it is convenient to

eliminate the noise present in the data base, by reducing

the dimensionality of data. Solutions to this problem are

to be found in the UL literature, and specifically among

the dimensionality reduction solutions.

• Pattern identification, grouping: The tasks belonging to

this class aim at identifying patterns, group of nodes

with similar characteristics, according to some kind of

criteria. An objective may be to apply to them similar

optimization approaches. Self-configuration use cases are

intuitive application for these issues. Solutions to these

problems can be translated into learning the set of classes

the data belongs to. UL literature offers solutions in the

area of clustering.

• Sequential decision problems for online parameter ad-

justment: This class of problems is extremely common

in the area of autonomous management, where we face

control decision problems to online adjust network pa-

rameters, with the objective to meet certain performance

targets. This kind of decision problems, where we learn

the most appropriate decision online, based on the re-

action of the environment to the actions the network is

taking, can be addressed through RL solutions. All self-

optimization use cases can be addressed through these

solutions, as well as COC problems.

In the rest of the section, we relate each class of NM

problem to the possible ML literature to solve it. The re-

view of ML literature provided in the following, is far from

being exhaustive. Many methods and techniques will not be

described, because the purpose is here to provide a useful

taxonomy to address NM and SON problems and to analyze

and understand the related literature using ML solutions. For

a deeper understanding of ML solutions, the reader is referred

to more specific literature.

A. Supervised Learning (SL)

This ML technique could be extremely useful when the

NM function to address requires estimation, prediction, clas-

sification of variables. SL is a ML technique which takes

training data (organized into an input vector (x) and a desired

output value (y)) to develop a predictive model, by inferring

a function f(x), returning the predicted output y. For that,

the construction of a dataset is needed. The dataset contains

training samples (rows), and features (columns), and is usually

divided into 2 sets. The training set, used to train the model,

and the test set, used to make sure that the predictions are

correct. The goal of the training model is to minimize the

error between the predictions and the actual values. Hence,

by applying ML, we aim to estimate how well a learning

algorithm generalizes beyond the samples in the training set.

The input space is represented by a n-dimensional input vector

x = (x(1), . . . , x(n))T ∈ Rn. Each dimension is an input

variable. In addition a training set involves m training samples

((x1, y1), . . . , (xm, ym)). Each sample consists of an input

vector xi, and a corresponding output yi. Hence x(j)i is the

value of the input variable x(j) in training sample i, and the

error is usually computed via |yi − yi|. The SL technique has

two main applications, classification and regression. On the

one hand, classification is applied when y, the output value we

try to predict is discrete, e.g., we want to predict if a cancer

is benign or malign, based on a dataset constructed based on

medical records, and collecting many features, e.g. tumour

size, age, uniformity of cell size, uniformity of cell shape. On

the other hand, a regression problem is applied when y is a

real number.

A huge amount of SL algorithms for classification can be

found in the literature, and a study to evaluate the performance

of some of them can be found in [91]. In the following we

briefly introduce the most common algorithms.

1) k-Nearest Neighbors (k-NN) can be used for classifica-

tion and regression. k-NN is a non-linear method where

the input consists of the k closest training samples in

the input space. The predicted output is the average of

the values of its k nearest neighbours. A commonly used

distance metric for continuous variables is the Euclidean

distance. The k-NN method has the advantage of being

easy to interpret, fast in training, and the amount of

parameter tuning is minimal. However, the accuracy of

the prediction is generally limited.

2) Generalized Linear Models (GLM). The linear model

describes a linear relationship between the output and

one or more input variables, and where the approxima-

tion function maps from xi to yi as follows,

yi = θ0 + θ1x(1)i + . . .+ θnx

(n)i (1)

where θi are the unknown parameters. The idea is

to choose θi so that yi minimizes the loss function.

Typically, we make the assumption that the samples in

each dataset are independent from each other, and that

the training set and testing set are identically distributed.

Note that if the relation is not linear, the model should

be generalized, in an attempt to capture this relationship

[92].

3) Naive Bayesian. The method is used for classification

and is based on Bayes theorem, i.e., calculating proba-

bilities based on the prior probability. The main task is

to classify new data points as they arrive. A NB classifier

9

assumes that all attributes are conditionally independent,

and is recommended when the dimensionality of the

input is high [93]. Since NB assumes independent vari-

ables, it only requires a small amount of training data

to estimate the means and variances of the variables.

4) Support Vector Machines (SVMs) can be used for classi-

fication and regression. SVMs are inspired by statistical

learning theory, which is a powerful tool for estimating

multidimensional functions [94], [95]. This method can

be formulated as a mathematical optimization problem,

which can be solved by known techniques. For this prob-

lem, given m training samples ((x1, y1), . . . , (xm, ym)),the goal is to learn the parameters of a function which

best fit the data. It samples hyperplanes. Thus, the

hyperplane with the main minimum distance from the

sample points is maintained. The sample points that

form margin are called support vectors and establish

the final model. This method in general shows high

accuracy in the prediction, and it can also behave very

well with non-linear problems when using appropriate

kernel methods. Also, when we cannot find a good linear

separator, kernel techniques are used to project data

points into a higher dimensional space where they can

become linearly separable. Hence the correct choice of

kernel parameters is crucial for obtaining good results.

In practice, this means that an exhaustive search must

be conducted on the parameter space, thus complicating

the task [96].

5) Artificial Neural Network (ANN) is a statistical learning

model inspired by the structure of a human brain,

where the interconnected nodes represent the neurons

producing appropriate responses. ANN supports both

classification and regression algorithms. The basic idea

is to efficiently train and validate a neural network. Then,

the trained network is used to make a prediction on the

test set. In this method the weights are the parameters

in charge of manipulating the data in the calculations.

Here, the interconnection pattern between the different

layers of neurons, the learning process for updating

the weights of the interconnections, and the activation

function that converts a neuron’s weighted input to its

output activation are the most important parameters to

be trained [97]. ANNs methods require parameters or

distribution models derived from the data set, and in

general they are also susceptible to over-fitting.

6) Decision Trees (DT) is a flow-chart model in which

each internal node represents a test on an attribute.

Each leaf node represents a response, and the branch

represents the outcome of the test [98]. DTs can be used

for classification and regression, and they have nuisance

parameters, such as the desired depth and number of

leaves in the tree [99]. Also, they do not require any

prior knowledge of the data, are robust (i.e., do not

suffer the curse of dimensionality as they focus on the

salient attributes) and work well on noisy data. However,

DTs are dependent on the coverage of the training

data as with many classifiers. Moreover, they are also

susceptible to over-fitting.

7) Hidden Markov Model (HMM) can be used for classifi-

cation, and also for other purposes. They can be used as

a Bayesian classification framework, with a probabilistic

model describing the data.

Methodologies have also been proposed to take the best

out of the available data, to boost the prediction performance.

Some of these methodologies are classified among the so

called Ensemble methods. Ensemble methods combine the

predictions of multiple learning algorithms to produce a final

prediction. This technique has been investigated in a huge va-

riety of works [100], [101]. A general method is sub-sampling

the training examples, where the most useful techniques are re-

ferred to as bagging and boosting [102]. Bagging manipulates

the training examples to generate multiple hypothesis. It runs

the learning algorithm several times, each one with different

subset of training samples. On the other hand, AdaBoost

maintains a set of weights over the original training set, and

adjusts these weights by increasing the weight of samples that

are misclassified, and decrease the weight of samples that are

correctly classified [103], [104].

B. Unsupervised Learning (UL)

This kind of learning can be extremely useful when the

NM function requires identifying anomalous behaviours, rec-

ognizing patterns or reducing the dimensionality of the data.

UL is a ML technique, which receives unlabelled input pat-

terns with the objective to find a pattern in it. In this case,

we let the computer learn by itself, without providing the

correct answer to the problem we want to solve. The goal

is to construct representation of inputs that can be used for

predicting future inputs without giving the algorithm the right

answer, as in turn we do in case of supervised learning [105].

The three most important families of algorithms are clustering,

dimensionality reduction and anomaly detection techniques.

There are many examples of UL applications in our daily

life, e.g., news.google.com, understanding genomics, organize

computer clusters, social network analysis, astronomical data

analysis, market segmentation, etc. In the context of SON, UL

algorithms are applied mainly on self-optimization and self-

healing use cases.

1) Clustering. This technique aims at identifying groups

of data to build representation of the input. The most

common methods to create clusters by grouping the

data are: non-overlapping, hierarchical and overlapping

clustering methods. K-means [106] and Self-organizing

Maps (SOMs) [107] methods belong to non-overlapping

clustering techniques. When the clusters at one level are

joined as clusters at the next level (cluster-tree), this is

referred in literature as a hierarchical clustering method

[108]. In case that an observation can exist in more than

one cluster simultaneously, this is known as overlap-

ping or fuzzy clustering. Fuzzy C-means and Gaussian

mixture models belong to this kind of technique [106],

[109]. Also HMM can be used for clustering This kind

of algorithms have been proposed in a wide range of

fields, such as, robotics, wireless systems, and routing

algorithms for mobile ad-hoc networks, among others.

10

2) Dimensionality Reduction. High-dimensional datasets

present many challenges. One of the problems is that, in

many cases, not all the measured variables are necessary

to understand the problem of interest. In the state of the

art we can find a huge amount of algorithms to predict

models with good performance from high-dimensional

data. However it is of interest for many problems to

reduce the dimension of the original data. For example,

in [110], [111], the authors face the problem of the huge

amount of potential features the system has as input,

and they suggest that the regression analysis has a better

performance in a reduced space. In this context, the most

common methods are: Feature Extraction (FE) and Fea-

ture Selection (FS) [112]. Both methods seek to reduce

the number of features in the dataset. FE methods do so

by creating new combinations of features (e.g. Principal

component analysis (PCA)), which project the data onto

a lower dimensional subspace by identifying correlated

features in the data distribution. They retain the Prin-

cipal Components (PCs) with the greatest variance and

discard all others to preserve maximum information and

retain minimal redundancy [113]. Correlation based FS

methods include and exclude features present in the data

without changing them. For example, Sparse Principal

Component Analysis (SPCA) extends the classic method

of PCA for the reduction of dimensionality of data by

adding sparsity constraint on the input features.

3) Anomaly Detection. Anomaly detection identifies events

that do not correspond to an expected pattern. By mod-

eling the most common behaviors, the machine selects

the set of unusual events [114]. Self healing is one of

the main functionality in which this kind of techniques

are applied, some examples are [115], [116]. The two

most common techniques are:

• Rule based systems: they are very similar to DTs,

but they are more flexible than DTs as new rules

may be added, without creating a conflict with the

existing ones [114].

• Pruning techniques: they aim at identifying outliers,

where there are errors in any combination of vari-

ables.

4) Latent Variable models. This kind of techniques allows

learning a model where some unseen variable helps

simplify and describe the data. An example is the non

negative matrix factorization.

C. Reinforcement Learning (RL)

The ML approaches under this category can be used to ad-

dress NM functions which require network parameter control.

Differently from the case of SL, RL aims to learn from interac-

tions how to achieve a certain goal. In many real applications

and in particular, in sequential decision and control problems,

it is not possible to provide an explicit supervision to the

training (i.e. the right answer to the problem). In these cases,

we can only provide a reward/cost function, which indicates

to the algorithm when it is doing well and when it is doing

poorly. RL has already been proven effective in many real

world applications, such as autonomous helicopters, network

routing, robot legged automation, etc. [117]–[119].

The learner or decision maker is called agent, and it interacts

continuously with the so-called environment. The agent selects

actions and the environment responds to those actions and

evolves into new situations. In particular, the environment

responds to the actions through rewards, i.e., numerical values

that the agent tries to maximize over time.

The agent has to exploit what it already knows in order to

obtain a positive reward, but it also has to explore in order to

take better actions in the future. Learning can be centralized in

a single agent or distributed across a multiple agents. In single

agent systems, ML approaches are capable of finding optimal

decision policies in dynamic scenarios with only one decision

maker. In multi agent systems, the distributed decisions are

made by multiple intelligent decision makers, and the optimal

solutions or equilibria are not always guaranteed [120].

The problem is then defined by means of a Markov De-

cision Process (MDP) {S,A, T ,R, γ}, where S, is the set

of possible states of the environment S = {s1, s2, . . . , sn},

A, is the set of possible actions A = {a1, a2, . . . , aq} that

each decision maker may choose, T (s′|s, a), is the transition

function denoting the probability of getting s′ giving an action

a in state s, R(s, a) is a reward function, which specifies the

expected immediate return obtained by executing action a in

state s, and 0 ≤ γ ≤ 1 is a discount factor, which gives

more importance to immediate rewards compared to rewards

obtained in the future [121].

The MDP represents the theoretical basis for the RL

framework [121]. At each time step, the agent implements a

mapping from states to probabilities of selecting each possible

action. This mapping is the agent’s policy.

The objective of each learning process is to find an optimal

policy π∗(s) ∈ A for each s, to maximize some cumulative

measure of the reward r received over time. Almost all RL

algorithms are based on estimating a so called value function,

which is a function of the states estimating how good it is

for an agent to be in a given state. For MDPs the state-

value function, denoted as Vπ(s), is the expected return when

starting in state s and following policy π thereafter. For more

information the reader is refereed to [121].

RL literature offers two approaches to solve MDPs. These

two approaches are: model-based and model-free.

1) Model-based. Dynamic Programming (DP) and Monte

Carlo (MC) methods fall into the category of model-

based approach.

a) DP is able to solve MDPs by relying on the

knowledge of the state transition probability be-

tween two states after executing a certain action.

DP is an algorithmic paradigm that solves a given

complex problem by breaking it into sub-problems

and stores the results of sub-problems to avoid

computing the same results again. DP algorithms

are based on update rules derived from the Bellman

equation. The first key component is known as

the policy evaluation process, according to which

a policy provides information about how much

reward is going to be received in the MDP. This

11

solution is used to build the first overall solution

by finding the optimal policy known as the policy

iteration process. Finally the value iteration makes

the value function better and better by applying

Bellman’s equation intuitively. DP is used to solve

problems such as, scheduling, graph algorithms,

bioinformatics, among others.

b) MC method only requires experience, i.e., sam-

ple sequences of states, actions and rewards. The

estimations are only updated after the episodes

conclude. Although their application on practical

cases is limited, they provide foundation for other

RL methods.

2) Model-free. Temporal Difference (TD) methods are

model free approaches to solve RL problems. TD learn-

ing is a combination of MC and DP ideas. It uses the

current estimate V πt of the value function instead of the

exact V π, as it happens in DP. If T is known, we can

solve the MDP through DP, otherwise we need to rely

on TD methods.

Some common examples of TD methods are: Q-

learning, Sarsa and Actor Critic (AC) [121]. TD methods

can be found in each SON functionality.

a) Q-learning and Sarsa are based on the estimation

of the state-action value function, Q(s, a). Learn-

ing is performed by iteratively updating the Q-

values, which represent the expert knowledge of

the agent, and have to be stored in a representation

mechanism. The most intuitive and common rep-

resentation mechanism is the lookup table, i.e., the

TD methods represent their Q-values in a Q-table,

whose dimension depends on the size of the state

and action sets. The difference between them is

that, Q-learning is an off-policy learner. This means

that, the agent will use the policy corresponding to

the best action in the next state, given the current

agent experience, whereas Sarsa is an on-policy

learner. On-policy learners evaluate the policy π, to

perform the decisions. This means that, the policy

followed by the agent to select its behaviour in

a given state is the same used to select the action

based on which it evaluates the followed behaviour.

b) AC methods have a separate memory structure to

represent the policy independently of the value

function. The policy structure is known as the

actor, since it is used to select the actions, while

the estimated value function is known as the critic.

The critic learns and critiques whatever policy is

currently being followed by the actor and takes the

form of a TD error δ, which is used to determine

if at was a good action or not. δ is a scalar signal,

which is the output of the critic and drives the

learning procedure. After each action selection, the

critic evaluates the new state to determine whether

things have gone better or worse than expected.

IV. MACHINE LEARNING ENABLED NETWORK

MANAGEMENT

As we have mentioned in the introduction of this work,

mobile networks constitute a huge source of data which could

be analyzed with proper tools, with the primary goal to make

more informed decisions when it comes to efficiently manage

the overall 4G or 5G network. In this context, ML is a great

opportunity due to its capability of providing insightful infor-

mation from the analysis of data already available to operators,

which can be used to make improvements or changes.

In this section we focus on how ML can specifically be

applied to SON and novel network management concepts.

First, we present all the relevant sources of information that

could be extracted from a mobile network. All these data are

available to operators, and may happen to be sensitive data

for the users’ privacy. However, some interesting data can

be derived from open databases or sniffed from unencrypted

control channels like the PDCCH. We will then discuss on

these options. Third, we will go through again the main SON

and network management functions and provide a classifi-

cation of the main inputs and outputs that we would need

available in the form of data, when designing an appropriate

ML algorithm to target the specific use case, and the KPI

indicators that we would need to monitor. Finally, we provide

an overview of SON and network management’s related work,

where ML techniques have been adopted, classifying this work

as a function of the targeted use case, the specific high level

problem to solve and the ML technique that the authors have

picked to address the problem.

A. Data generated by mobile cellular networks

As we observed in [18] a huge amount of data is cur-

rently already generated in mobile networks during normal

operations by control and management functions. This kind

of data can be exploited to find patterns and extract useful

information from them. This allows to take more informed

decisions to effectively manage network performance. Some

examples of the different sources information generated by

mobile networks, together with the kind of usage currently

provided by operators, and related references of interest, is

detailed in Table II.

1) Charging Data Records (CDR). They are defined in

[122] and provide a comprehensive set of statistics at the

service, bearer and IP Multimedia System (IMS) levels.

These records are typically stored for offline processing

by the operator. The granularity of this information in

the time domain is however quite coarse, as records

are generated in correspondence with high-level service

events (e.g., start of a call).

2) Performance management functionality. This data source

[123] [36] provides data regarding the network per-

formance and it covers, among others, aspects of the

performance of the radio access network, such as, radio

resource control and utilization, performance of the

various bearers (both on the radio part and in the back-

haul), idle and connected mode mobility.

12

3) Minimization of Drive Tests (MDT). The data extracted

from this source refers to the radio measurements of both

idle and connected mode mobility, coverage items, such

as, power measurements and radio link failure events,

and can be associated with position information of the

UE performing the measurement. More information on

these data has already been provided also in section II-F.

4) E-UTRA Control plane protocols and interfaces, such

as Radio Resource Control (RRC), S1-AP, X2-AP pro-

tocols, are another huge source of information, espe-

cially concerning aspects, such as cell coverage, user

connectivity, mobility in idle and connected mode, inter-

cell interference, resource management, load balancing,

among others.

5) Data plane traffic flow statistics, also are a huge source

of information, which can be gathered at various points

of the network, like the eNB, or the PDN Gateway

(PGW) and Serving Gateway (SGW). The Internet Pro-

tocol Flow Information Export (IPFIX) is an example of

standardized format to exchange this kind of statistics

[124].

All these data are available to the network operators, but

in most cases this is not made available to the academic

community due to privacy issues and network operators’

interests. There are some exceptions, like the Data for Devel-

opment (D4D) initiative from the Orange group [126], which

made available anonymous data extracted from the Senegal’s

network to research laboratories. However these data are in

general aggregated and do not allow deep insight into the

operator’s network.

This lack of data represents a great limitation for the

advancement of the ML based network management research.

However, some network data can be derived in other ways.

Some databases are available, providing a huge insight in

mobile network operators. Some examples are listed in the

following, together with information that can be extracted from

them.

• opencellid: It contains information on specific cells, such

as: network type (GSM, UMTS, LTE), Mobile Country

Code (MCC), Mobile Network Code (MNC), Location

Area Code (LAC) for GSM and UMTS, Tracking Area

Code (TAC) for LTE, Cell ID for (CID) for GSM and LTE

networks, Primary Scrambling Code (PSC) for UMTS

networks, Physical Cell ID (PCI) for LTE networks,

longitude and latitude in degrees, estimates of range in

meters, total number of measurements collected from the

tower, defines if the coordinates of the cell tower are exact

or estimated, information of the date when the cell tower

was first added to the data base and updated, average

signal strength from all measurements received from the

cell in dBm, or as defined in [127]. This data base also

receives funding from important vendors like Qualcomm

[128] and offers some formula of free access to portion

of data for academic purposes.

• opensignal: It offers information on achievable data rates,

latency and availability, per operators, but not information

per cell tower [129].

• antenasgsm: It offers information on maps and positions

of cells, with added information on the operator and the

assigned bandwidth [130].

• Google geolocations API: It allows queries based on the

cell ID to get cell related information and WiFi Access

Points (AP), such as latitude and longitude [131].

The information provided by these databases is precious, but

still does not give sufficient insight on the behaviour of the

network, and mainly offers an overview of the coverage pro-

vided by the single operators. To get more information, still we

can do something more and access directly to the unencrypted

PDCCH and extract information exchanged between the users

and the associated eNB. In particular, it is possible to build a

sniffer, as the one described in [132], from which to collect raw

communication traces exchanged by the users and the associ-

ated eNodeB. This allows to have access not only to aggregate

base station statistics, but also to more valuable information

derived from the radio protocols, such as the resource block

allocation and the link adaptation mechanism of the system.

In particular, the OWL sniffer [132] is an online decoder

of the LTE control channel, which uses a Software Defined

Radio (SDR) to send the raw LTE signal to a PC running the

decoding software. This open-source software is capable of

reliably logging the LTE downlink control information (DCI)

broadcasted by base stations. In fact, LTE uses an unencrypted

control channel to assign network resources to users for both

downlink and uplink communications. Resources are assigned

to devices through their radio network temporary identifiers

(RNTIs), every millisecond, specifying the number of resource

blocks (RBs) and the modulation and coding scheme (MCS)

to be used. There are works in literature using this sniffer to

collect and analyze traces from different European cities [133].

Finally, let us review the main SON use cases in Table

III, by analyzing the main input information that their design

would require, in terms of data, together with the main

identified output actions and meaningful associated KPIs.

B. Overview of ML based Network management’s relevant

literature

This section reviews SON and Network management’s

recent work in the area of ML. We will go through each main

function and use case and review significant literature and ML

approach that has been used to approach the problem. Table IV

summarizes the main works in this area and classifies them

per 3GPP use case, technique and specific algorithm adopted

by the authors.

1) Use case: Indicates the 3GPP targeted use case.

2) Reference: Indicates the reference of the related work.

3) Technique: Indicates the applied ML method (Super-

vised Learning, Unsupervised Learning, Reinforcement

Learning).

4) Problem: Indicates the general problem to solve.

5) Algorithms: Indicates the specific ML algorithm applied

to the data (see TableIV).

1) Mobility Load Balancing: The literature offers some

examples of application of ML techniques to the MLB use

case. The majority of applications fall in the area of RL, as

13

TABLE II: Information elements relevant for ML enabled SONs

Source Data Usage TS

Charging Data Records(CDR)

Includes statistics at the service, bearer and IP Multimedia Subsystem(IMS) levels.

These records are typically stored, but onlyused by customer service. The network opera-tion departments typically do not leverage thisinformation and do not have access to it, as muchas customer service does not leverage networkmanagement data.

TS 32.298 [122]

Performance management(data on network perfor-mance)

It covers long-term network operation functionalities, such as Fault,Configuration, Accounting, Performance and Security management(FCAPS), as well as customer and terminal management. An exam-ple is that defined for Operations, Administration, and Management(OAM), which consists of aggregated statistics on network perfor-mance, such as number of active users, active bearers, successful/failedhandover events, etc. per BS, as well as information gathered by meansof active probing.

The data is currently mostly used for faultidentification, e.g., triggering alarms when someperformance indicator passes some threshold, sothat an engineer can investigate and fix the prob-lem. Typically, the only automatic use of thisinfo is threshold-based triggering, which can bedone with very low computational complexity.

TS32.401 [123],TS32.425 [36]

Minimization of DriveTests (MDT)

Radio measurements for coverage, capacity, mobility optimization,QoS optimization/verification

This data is used for identified use cases suchas coverage, mobility and capacity optimization,and QoS verification

TS37.320 [125]

E-UTRA Control planeprotocols and interfaces

Control information related to regular short-term network operation,covering functionalities such as call/session set-up, release and main-tenance, security, QoS, idle and connected mode mobility, and radioresource control.

A This information is normally discarded afternetwork operation purposes have been fulfilled.Some data can be gathered via tracing function-ality or used by SON algorithms which normallydiscards the information after usage

TS36.331 [48],TS36.413 [39],TS36.423 [40]

TABLE III: SON inputs, outputs and KPIs

SON function Inputs Output actions KPIs

Mobility Load Balancing (MLB) X2 resource status and load estimation in-formation.

Tuning the CIO, i.e. offsets of serving and neighbourcells to satisfy handover conditions.

Improved QoS and capacity

Mobility Robustness/HandoverOptimisation (MRO)

S1AP and X2AP handover requests, han-dover reports, RLF reports and indications.

A3 offsets, TTT, L1 and L3 filter coefficients, inconnected mode, and Qoffset in Idle mode.

Minimized call drops, RLFs andping pong effects.

Coverage and Capacity Optimiza-tion (CCO)

UE measurements Transmission power, pilot power, antenna parame-ters, coordinated Almost Blank Subframes (ABS)

Maximized coverage and cell andedge throughput

Inter-Cell Interference Coordina-tion (ICIC)

HII, RNTP, OI, UE Measurements. Transmission power, pilot power, antenna parame-ters, coordinated ABS

Minimized Intercell interference.

Cell Outage Compensation (COC) UE Measurements. Transmission power, antenna parameters of neigh-bouring cells

Minimized outage.

Energy Saving (ES) Resource status, UE Measurements. Switch ON and OFF policies Minimized energy consumption.

the main problem to solve is a sequential decision problem

about how to set configuration parameters, which optimize

network performance and user experience. An example of a

RL application for MLB use case can be found in [135].

Here the authors present a distributed Q-learning approach

that learns for each load state the best MLB action to take,

while also minimizing the degradation in HO metrics. Another

option to take advantage also of fuzzy logic capabilities of

dealing with heterogeneous sources of information is provided

in [136], where fuzzy logic is combined with Q-learning in

order to target the load balancing problem. For similar reasons,

fuzzy logic is also proposed in [137] to enhance the network

performance by tuning HO parameters at the adjacent cells.

Approaches incorporating fuzzy logic with RL capabilities

have the advantage to capture the uncertainty existing in real

world complex scenarios, while schemes considering only

learning approaches may be limited by the fixed variable

definition. When combining fuzzy logic with RL, also the

subjectivity with which the fuzzy variable may be defined

is overcome by the adjusting capabilities of the learning.

Alternatively, a centralized solution is approached in [138],

where a central server in the cellular network determines all

HO margins among cells by means of a dynamic programming

approach. Besides RL, also clustering schemes have been

proposed in this area, to group cells with similar characteristics

and provide for them similar configuration parameters [139].

Considering clustering in large realistic scenarios is an added

value to reduce computational complexity and take advantage

of what is learnt in other regions of the network where we

observe similar environment characteristics.

2) Mobility Robustness Optimization: Also for the case of

MRO, we find in literature different solutions based on RL to

solve a control decision problem. In [141], [142], the authors

focus on the optimization of the users’ experience and of

the HO performance. In [141] the authors take advantage of

the Q-learning approach to effectively reduce the call drop

rates, whereas in [142], unlike other solutions that assume a

general constant mobility, the authors adjust the HO settings in

response to the mobility changes in the network by means of

a distributive cooperative Q-learning. Differently from [141],

[142], in [143], the authors take advantage also of fuzzy logic

capabilities. These solutions are based on control optimiza-

tion of HO parameters through RL, so they propose similar

solutions to those found in the literature of MLB. In this

case we can do the same considerations about the advantages

of considering fuzzy logic in order to gain in flexibility in

the uncertain and complex real network context. Different

approaches in turn, address the problem by identifying suc-

cessful HO events, through solutions based on unsupervised

learning. In particular, the works of [144] and [145] propose an

approach to HO management based on UL and SOM analysis.

The idea is to exploit the experience gained from the analysis

14

TABLE IV: Related work

Reference ML technique Problem Algorithm

Self-configuration

PCI [134] UL Planning Clustering

Self-optimization

MLB [135] RL Control optimization Q-learning[136] RL Control optimization Q-learning[137] RL Control optimization Fuzzy Q-learning[138] RL Control optimization Dynamic Programming[139] UL Grouping K-means clustering[140] SL Prediction Multivariate polynomial regression

MRO [141] RL Control optimization Q-learning[142] RL Control optimization Q-learning[143] RL Control optimization Fuzzy control[144], [145] UL Pattern identification SOM[146] UL Prediction Semi-Markov model[147]–[150] SL Prediction ANN

CCO [151] RL Control optimization Fuzzy Q-learning[152] RL Control optimization Fuzzy Q-learning[153] RL Control optimization Fuzzy Q-Learning[154] UL, RL Control optimization Fuzzy ANN/Q-learning

ICIC [155] RL Control optimization Q-learning[156] RL Control optimization Fuzzy Q-learning[157], [158] RL Control optimization Q-learning

ES [159] RL Control optimization Q-learning[160] UL Decision making Fuzzy logic[161], [162] UL Grouping, pattern identification Clustering

Self-healing

COC [163] RL Control optimization Actor Critic[164] RL Control optimization Actor-Critic[165] SL Control optimization Fuzzy logic

COD [166] UL Anomaly detection Diffusion Maps[167] SL Anomaly detection Fuzzy logic[168] SL/UL Diagnosis Naive Bayesian[169] SL Anomaly detection SVM, Ensemble methods[164] SL/UL Anomaly detection k-NN, local-outlier-factor[170] UL Grouping, pattern identification Hidden Markov Model[171]–[173] SL Fault Detection k-NN[174], [175] SL Diagnosis Naive Bayesian

Self-coordination

[176] SL Classification Decision Trees[177] RL Control optimization Actor Critic[178] RL Control optimization Q-learning[179] RL Control optimization Actor Critic

Minimization Drive Tests

[180], [181] SL Verification/estimation Linear correlation[110] SL Prediction Regression models[182] SL/UL Prediction/curse of dimensionality Regression models/Dimensionality reduction[111], [183] SL Prediction Bagged-SVM/Dimensionality reduction

Core Networks[184] SL Prediction Adaboost, SVM

of data of the network based on the angle of arrival and the

received signal strength of the user, to learn specific locations

where HOs have occurred and decide whether to allow or

forbid certain handovers to enhance the network performance.

The solutions enable self-tuning of HO parameters to learn

optimal parameters’ adaptation policies. Similarly, in [146] the

authors exploit the huge amount of information generated in

the network to predict user traffic distribution. In particular,

they take advantage of semi-Markov model for spatiotemporal

mobility prediction in cellular networks. Finally, the works in

[147]–[150], propose schemes to make predictions about UE’s

mobility, which allows to anticipate smart HO decisions.

3) Coverage and Capacity Optimization: In case of CCO,

different approaches in literature focus on RL solutions based

on continuous interactions with the environment, oriented to

online adjusting antenna tilts and transmission power levels

through TD learning approaches. In [151] and [152] a fuzzy Q-

learning approach to optimize the complex wireless network,

by learning the optimal antenna tilt control policy has been

proposed, and a similar approach is followed also in [153]

and [154]. In addition, they also propose to combine fuzzy

logic with Q-learning, in order to deal with continuous input

15

and output variables. [153] also proposes a central control

mechanism, which is responsible to initiate and terminate the

learning optimization process of every learning agent deployed

in each eNB. Finally, [154] innovates with respect to other

approaches since in order to adjust the antenna tilt and trans-

mission power parameters, it considers the load distribution

of the different cells involved in the optimization process, and

introduces novel mechanisms to facilitate cooperative learning

among the different SON entities.

4) Inter-cell Interference Coordination: Similarly to the

CCO case, ML has been proposed in the literature of ICIC use

case as a valid solution, where RL is the principle used tool,

with special emphasis to TD methods, in order to target the

optimization of control parameters. Several works target the

problem to minimize the interference among cells by using the

most common TD learning method, Q-learning [155]–[158].

The work in [155] is related to control inter-cell interference in

a heterogeneous femto-macro network. The work combines in-

formation handled by the multi-user scheduling with decisions

taken by a learning agent based on Q-learning, which tries to

control the cross-tier interference per resource block. [156]

proposes a distributed solution for ICIC in OFDMA networks

based on a Fuzzy Q-learning implementation. The proposed

solution achieves joint improvement for all users, i.e., the

improvements of users with bad quality does not come at the

expense of users with good quality. Moreover, a decentralised

Q-learning framework for interference management in small

cells is proposed in [157]. The authors focus on a use case in

which the small cell networks aim to mitigate the interference

caused to the macro-cell network, while maximizing their own

spectral efficiencies. Finally, in [158] also a decentralized Q-

learning approach for interference management is presented.

The goal is to improve the systems performance of a macro-

cellular network overlaid by femto-cells. In order to improve

the time of convergence, a mitigation approach has been

introduced, allowing them to have significant gains in terms

of throughput for both, macro and femto users. Interesting

trade-offs can be studied to compare centralized vs. distributed

solutions. In the novel context of small cells distributed solu-

tions to interference management are to be preferred over more

complex centralized solutions, but convergence and instability

approaches may appear to affect the TD learning schemes,

compromising system performances [155].

5) Energy Savings: Energy savings schemes for wireless

cellular systems have been proposed in the past, enabling cells

to go into a sleep mode, in which they consume a reduced

amount of energy. In order to reduce the energy consumption

of the eNBs, we can found several works related to ML

techniques. An example of that can be found in [159], where

the authors take advantage of RL to propose a decentralized Q-

learning approach to allow energy savings by learning a policy

by the iterations with the environment taking into account

different aspects over time, such as the daily solar irradiation.

Also, in [160], the authors switch off some underutilized cells

during off peak hours. The proposed approach optimizes the

number of base stations in dense LTE pico cell deployments

in order to maximize the energy saving. For the purpose,

they use a combination of Fuzzy Logic, Grey Relational

Analysis and Analytic Hierarchy Process tools to trigger the

switch off actions, and jointly consider multiple decision

inputs for each cell. This last work uses smart decision theory

approaches, which though are not able to take advantage of

the previous decisions made in the same environment, as in

turn does the work proposed in [159], as a result of the TD

learning approach. This allows that the work in [159] offers

a more solid solution, considering also past information in

the decision. Also for HetNets, we find several works, such

as, [161], [162], where the authors take advantage of KPIss

available in the network for the construction of different kind

of databases to analyse the potential gains that can be achieved

in clustered small cell deployments.

6) Cell Outage Compensation: The literature already offers

different works targeting the problem of COC. For this use

case RL has been proven as a valid solution since it is a

continuous decision making/control problem. In this context

a contribution in the area of self-healing has been presented

in [163], [164], where the authors present a complete solution

for the automatic mitigation of the degradation effect of the

outage by appropriately adjusting suitable radio parameters

of the surrounding cells. The solution consists of optimizing

the coverage and capacity of the identified outage zone, by

adjusting the gain of the antenna due to the electrical tilt

and the downlink transmission power of the surrounding

eNBs. To implement this approach, the authors propose a RL

based on actor-critic theory to take advantage of its capability

of making online decisions at each eNB, and of providing

decisions adapting to the evolution of the scenario in terms

of mobility of users, shadowing, etc., and of the decisions

made by the surrounding nodes to solve the same problem.

A COC contribution also based on ML is targeted in [165],

where fuzzy logic is proposed as the driving techniques to

fill a coverage gap. The authors show performance gains by

using different parameters, such as, the power transmission,

the antenna tilt, and a combination of the two schemes. These

two works are compared in [163] and the work in [163] is

proven superior thanks to the ability to learn from the past

experience introduced by the RL actor-critic approach.

7) Cell Outage Detection: As we already mentioned, COD

aims to autonomously detect cells that are not operating

properly due to possible failures. For this kind of problem,

anomaly detection algorithms offer an interesting solution

that allows to identify outliers measurements, which can be

highlighting a hidden problem in the network. Proposals of

solutions for this problem can be found in [166] and [167]. In

particular, [166] presents a solution based on diffusion maps,

by means of clustering schemes, capable of detecting anoma-

lous behaviours generated by a sleeping cell. [167] presents

a solution based on fuzzy logic for the automatic diagnosis

of a troubleshooting system. In order to determine if there is

a failure, the authors propose a controller, which receives as

an inputs a set of representative KPIs. A similar approach is

presented by [168], where the authors present an automated

diagnosis model for Universal Mobile Telecommunications

System (UMTS) networks based on Naive Bayesian classifier,

and where the model uses both network simulator and real

UMTS network measurements. In the context of this king of

16

classifiers , the works in [174], [175], also take advantage of

NB for automated diagnosis based on different inputs network

performances. The work in [169] addresses both the case

of outage and the one where in turn the cell can provide

a certain level of service, which though does not allow to

fulfil the expected UEs requirements. The approach relies on

ensemble methods to train KPIs extracted by human operators

to make informed decisions. In [185], the authors consider

large data sets to identify anomaly behaving base station. They

proposed an algorithm consisting of preprocessing, detection

and analysis phases. The results show that by using dimension-

ality reduction and anomaly detection techniques irregularly

behaving base stations can be detected in a self-organized

manner. In [164] data gathered through MDT reports is used

for anomaly detection purposes. Furthermore, the works of

[171]–[173] take advantage of k-NN algorithm to propose a

self-healing solution, in particular to tackle the fault detection

domain. Finally, in [170], the authors consider a HetNet and

they take advantage of HMM to automatically capture the

dynamic’s of four different states and probabilistically estimate

if there exist a possible failure.

8) SON Conflicts Coordination: As the deployment of

stand-alone SON functions is increasing, the number of con-

flicts and dependencies between them also increases. Hence,

an entity has been proposed for the coordination of this kind

of conflicts. In this context, current literature includes several

works based on ML. In [176] the authors focus on the classi-

fication of potential SON conflicts and on discussing the valid

tools and procedures to implement a solid self-coordination

framework. Q-learning, as a RL method, has been proposed

in [177] to take advantage of experience gained in past

decisions, in order to reduce the uncertainty associated with

the impact of the SON coordinator decisions when picking an

action over another to resolve conflicts. In [178], the authors

use Q-learning to deal with the conflict resolution between

two SON instances. Decision trees have been proposed in

[186] to properly adjust Remote Electrical Tilt (RET) and

transmission power. Additionally, in [179] the authors provide

a functional architecture that can be used to deal with the

conflicts generated by the concurrent execution of multiple

SON functions. They show that the proposed approach is

general enough to model all the SON functions and their

derived conflicts. First they introduce these SON functions

in the context of the general SON architecture, together with

high-level examples of how they may interfere. Second, they

define the state and action spaces of the global MDP that

models the self-optimization procedure of the overall RAN

segment. Finally, they show that the global self-optimization

problem can be decomposed onto as many Markov decision

sub-processs (subMDPs) as SON functions.

9) Minimization of Drive Tests: The great majority of

literature using the MDT functionality to target MDT use

cases, takes advantage of supervised and unsupervised learning

techniques to provide different solutions for the different use

cases. An example of that can be observed in [180], [181],

where the authors address the QoS estimation by selecting

different KPIs and correlating them with common nodes

measurements, to establish whether a UE is satisfied with

the received QoS. A similar objective is targeted in [110],

however, differently from the previous works, here the authors

focus on multi layer heterogeneous networks, so in a more

complex and realistic scenario than the traditional macrocell

one. In particular, they present an approach, based on regres-

sion models, which allows to predict QoS in heterogeneous

networks for UEs, independently of the physical location of

the UE. This work is extended in [182] by taking into account

the most promising regression models, but also analysing

dimensional reduction techniques. By doing PCA/SPCA on the

input features, and promoting solutions in which only a small

number of input features capture most of the variance, the

number of random variables under consideration is reduced.

Based on previous results, in [111], [183] the same authors

define a methodology to build a tool for smart and efficient

network planning, based on QoS prediction derived by proper

data analysis of UE measurements in the network.

Moreover, the work in [187] presents a system based on

a fuzzy logic controller to improve network performances by

adjusting antenna tilts values in a LTE system. Differently

from previous works, the authors consider the use of call

traces to identify the level of coverage, overshooting and

overlapping problems, which are the inputs to the algorithm.

Also, in [188], the same authors take advantage of connection

traces (signal strength, traffic, and resource utilization mea-

surements) to improve the network infrastructure in terms of

spectral efficiency. The proposed method is designed to be

integrated in commercial network planning tool. Finally, in

[189] the authors take advantage of the MDT measurements

to build a Radio Environment Map (REM) by applying spatial

interpolation techniques (Bayesian kriging). The REM (Radio

Environmental Map) is then used to detect coverage holes and

predict the shape of those areas.

10) Core Networks: As we already mentioned in section

II, the operational aspects of core networks elements can be

enhanced through, for example, the automatic configuration of

the neighbour cell relations function. In this regard, the idea of

applying ML to this function is not new. In [184] the authors

study the benefits of using ML to root-cause analysis of session

drops, as well as drop prediction for individual sessions. They

present an offline Adaboost and SVM method to create a

predictor, which is in charge of eliminating/mitigating the

session drops by using real LTE data.

11) Virtualized and Software Define Networks: Also when

we go beyond the RAN and we focus on the network in

general, ML concepts have already been proposed in different

works to build cognitive based techniques to operate the

network. An example of these proposals is well summarized

by [190]. In this work, a Knowledge Plane is advocated, which

would bring many advantages to the networks in terms how the

network is operated, automated, optimized and troubleshooted.

Conceptually this vision is aligned with different others pro-

posals in other areas, such as the black-box optimization [191],

the autonomic self-x architectures [192], or the work presented

in [193]. In this context, the work in [194] analyzes the reasons

why the vision proposed in [190] has still not been brought to

reality, and the main reason that they find is in the challenges

that appear when it comes to autonomously manage a network

17

in a distributed fashion. In particular, the work argues that

the emerging trend of centralization in control brought by the

novel SDN vision, will significantly reduce this complexity

and favour the realization of the ML vision in the network.

As a result, in [194] some initial experimental results based

on the vision defined in [190] are brought into reality in the

context of a SDN based architecture. Further work in this

area is carried put in the context of different European H2020

projects [6]. The work in [195] presents a novel cognitive

management architecture that manages multiple use cases,

like the Service Level Agreement (SLA) and the Mobility

Quality Predictor. Both use cases are tackled using machine

learning approaches, the Long Short Term Memory, and a per

user bandwidth predictor. The work in [196] implements SLA

through ML approaches. It uses an ANN for evaluation of

cognitive SLA enforcement of networking services involving

Virtualized Network Functions and SDN controllers.

V. CHALLENGES FOR FUTURE WORKS

In this section, we focus on some open challenges that still

need to be addressed when it comes to making ML based

network management a reality.

A. Real data

It is possible to find databases related to signals and cov-

erage data [126], [129], by using/designing applications that

collect information such as Reference Symbol Received Power

(RSRP) and Reference Symbol Received Quality (RSRQ).

However, it is not easy to find contributions analysing real

network management data. Some work can be found in the

context of 3G networks, but currently, in the context of

4G networks it is very hard to find works considering real

data [197], [198]. These works though do not take into

account the data analysing them through ML techniques, to

extract experience from them. We consider that it is extremely

important for this research line to get to the next level,

to get access to operator’s network data. An alternative to

real data, could be to sniff data from unencrypted LTE

control channels, as we have shown in [133] or to use a

high-fidelity network simulator ns-3 LTE/LTE-EPC Network

Simulator (LENA) module, to generate realistic data [199].

This simulator has been built around industrial Application

Programming Interface (API) defined by the small cell forum

and offers high-fidelity models from Media Access Control

(MAC) to application layers. It has also been designed with

the requirement to simulate tens of eNBs and hundreds of UEs,

and to specifically test Radio Resource Management (RRM)

and SON algorithms. Consequently it could be a very useful

tool to build realistic scenarios based on information available

in public databases, generate data to analyse, build algorithms

based on this analysis and close the loop on the simulator to

test the designed algorithms. In this context, it is also hard to

find contributions where ML approaches are used not only

in network simulators, but in real networking products. In

general, it seems vendors are reluctant to test algorithms whose

behaviour is not predictable. An important research line is then

how to find or generate meaningful network data, and find

patterns in them to understand aspects that should be optimized

in the network.

This research line, additionally, faces important privacy and

confidentiality issues. It is important to ensure that the data

that is used is properly anonymized. As mentioned in section

IV, data come from different sources of the network, but

can also be offered by third parties, e.g., data generated by

the user, open data, sensor data, among others. Therefore, to

come up with a unified privacy policy is extremely challenging

at security and privacy levels, due to the variety and the

granularity of the data. If we add to this the speed at which data

are created and need to be analysed, the security challenges

are huge. In this context, big data is changing the security

analytics, where robust and scalable privacy preserving mining

algorithms are critical to ensure that the most sensitive private

data is secure. As a result, privacy-preserving data mining

is a challenging research line that has to be investigated.

In particular, in order to guarantee privacy protection, it

is important to define the privacy requirements taking into

account the lifecycle of the analytics. For example, in the

data collection phase, it is important to identify the personal

data needed for processing. The idea is to extract only the

needed data for the specific purpose. Aggregated information

can also be used instead of personal data. In this context, one

of the most relevant techniques is anonymization, which is

the process of modifying personal data in such a way that no

identification is possible. Regarding the data analysis phase,

different privacy models are available in the context of big data

analytics, where two of the most important families are: K-

anonymity and differential privacy. A review in more detail of

the aforementioned methods can be found in [200]. Moreover,

in order to protect personal data in databases (data sotrage

phase), encryption is a fundamental security technique, which

transforms data in a way that only authorized parties can read

it. For more information, the reader is referred to [200], [201].

B. Big Data and Deep Learning

Deep learning is a new trend in ML that allows computer

systems to improve with experience and data. It achieves great

power and flexibility to operate in complicated real-world

environments, by learning to represent the world through a

nested hierarchy of concepts. The ML algorithms that we

have reviewed in this paper have a strong dependency on the

features on which the algorithms are applied. Based on that,

much effort has been devoted to design ML algorithms that

yield to useful representations. This is known as representation

learning, and deep learning is one way of learning representa-

tions [202], [203]. The main representation of deep learning is

through a Multilayer Perceptron, which is a multilayer neural

network function mapping some sets of input values to output

values. Each layer of this representation learns a hierarchy

of the output. Deep Learning has the ability to do successful

training from the bottom layers to the higher ones. This is

done by applying computational models that are composed

of multiple levels of representation and abstraction that help

make sense of data.

Historically, DL has become more useful as the amount

of training data has started increasing (big data). Also, the

18

research on deep learning has benefited from the increase of

computer infrastructure at both hardware and software levels.

All this has made that deep learning has solved increasingly

complicated applications with increasing time and accuracy.

The potential of these improved techniques in the area of

NM, in case the big data associated to the management of

the complex 4G, 5G network ecosystem is available, is still to

be evaluated and open for research.

C. Theoretical research

With respect to online control decision problems that allow

to continuously take RRM/SON decisions, we are aware of

some approaches, which take advantage of reinforcement

learning to solve this problem. The current approach is to

use single agent algorithms and extend them to multi-agent

settings. However, this kind of algorithms require a consider-

able amount of time before finding a solution, and it increases

with the state and action spaces. So, reinforcement learning

approaches dealing with this issues have to be investigated.

Moreover, no proof of convergence is available demonstrating

that this approach actually reaches meaningful conclusions.

Even though ML literature offers different algorithms that

can find interesting solutions (e.g. NashQ [204]), the space

of possible solutions is so big that this kind of approaches

is not feasible to be used in a realistic network where the

time constraints of RRM/SON problem have to be met. So,

more research in the area of multi-agent systems, which are

also compatible with real network requirements need to be

investigated.

In the context of data analytics, it is well known that

the analysis of the data requires a substantial amount of

”black art”, and consequently it requires the availability in

the research groups of multi-disciplinary researcher profiles

knowledgeable of information technology, computer science,

and telecom engineering, to properly optimize the network

accordingly. In this context, ML trends, like deep learning can

be very useful, however little work can be found applying these

new promising techniques to network management [205].

D. Network management of multi-technologies networks and

of future New Radio

Autonomous network management of multi-technology net-

works, where heterogeneous networks including different Ra-

dio Access Technologies (RAT), or different layers of the

network are coexisting, e.g., Wi-Fi, mmWave, mobile network

layer, transport layer, among others, is still immature. How-

ever, these scenarios will tend to emerge with the advent of the

unlicensed spectrum paradigm and with technologies such as

LAA or New Radio (NR), the new radio access technology

for 5G, which is currently under definition in 3GPP. NR,

in particular, will be defined to work over a wide range of

spectrum opportunities, ranging from sub 6 GHz and up to

mmWave spectrum, and under multiple spectrum paradigms,

such as licensed, unlicensed and shared. The opportunities

of autonomous network management in this area are huge.

ML has still not been exploited to handle these networks

with intelligence and self-awareness. In particular, the man-

agement of densified and heterogeneous, in both technologies

and layers, architectures, requires the evolution of complex

SON concepts, which have traditionally been designed and

standardized for LTE based networks. Also, self-organization

in the context of NR technology is still to be completely

defined. Before reaching this vision, multiple challenges need

to be addressed, e.g. the self-coordination problem and the

solution of conflicts among SON functions executed in dif-

ferent nodes, or networks, which put the network at risk of

instability, or the most appropriate location of SON functions

and algorithms, to solve properly the distributed vs. central-

ized SON implementation issue. Many aspects have to be

considered when locating and designing a SON function, e.g.

response time, complexity, size of databases, computational

capability of nodes, etc. Centralized (i.e. a large number of

cells is involved), distributed (approx. 2 cells are involved,

coordinating through X2) and local (only one cell is involved)

implementations of SON functions have been proposed. No

architecture can be claimed superior to the other. The growing

complexity, dynamicity, and heterogeneity of 5G networks

will substantially increase the number of scenarios to solve.

So, there is the need for exploiting their complementarity by

virtualizing and dynamically deploying them.

E. Network management of novel softwarized and virtualized

architectures

To benefit of all the opportunities offered by centralized,

distributed and local implementations, and towards the need of

virtualizing resources in order to reduce network costs, while

meeting the stringent new service verticals’ requirements,

there is the need to further study autonomous NFV and

SDN architecture, where end-to-end SON functions, aimed

at tackling the main radio access and backhauling challenges

of extremely dense deployments, are virtualized and run

over generic purpose hardware. This infrastructure is to be

managed by an orchestrator entity (in coordination with the

corresponding virtual network function and virtual infrastruc-

ture managers), as proposed in European Telecommunications

Standards Institute (ETSI) architecture. This orchestrator or

SDN controller is the brain of the network and needs the ability

to adapt to ever-changing conditions. The network should not

only react to failures, but adapt to the demand, predict it, based

on data analytics, and facilitate in this way the task of the net-

work management. Research on deep reinforcement learning

implementations of the orchestrator will allow the controller to

self-learn after every decision. Automation will require also all

the advancements of the Information Technology sector, with

increased computational capacity, more CPUs and memory

space. However, future orchestrators will need to handle a

huge amount of data and learn from them through novel deep

learning approaches, the smart network management decisions.

This research line is still highly immature and requires a lot

of efforts.

VI. CONCLUSIONS

In this work we have motivated the need for ML to be

considered as a crucial and inevitable tool in order to address

19

automation, self-awareness and self-organization in current

and future mobile networks. The SON features have been

considered fundamental in LTE definition and have been

introduced in this technology since its very beginning in

Release 8. We believe that this need of automation will be

further enhanced with the expected complexity that future

5G network management will have to handle. On top of

that, we have shown that current cellular networks, already

generate a huge amount of data that if properly stored and

managed could bring new insights in how the networks work

and offer new challenges for improving network management

taking into account the experience that can be gained from

these data. We have reviewed the main taxonomy of machine

learning and the novel trends that could make this exploitation

of data to gain insight of the network a reality. Also, we

have discussed open data options, as much as alternatives to

get data from the networks, which otherwise are not made

available to the academic community. With this motivations

in mind, we have started by reviewing the main concepts and

taxonomy of SON, network management and ML, and we have

reviewed significant academic literature in the area of network

management, focusing only on solutions based on ML. The

work has reviewed the status of this exciting research line,

while at the same time highlighting open challenges that we

need to deal with in order to make future autonomous network

management a reality.

VII. COMPETING INTERESTS

The authors declare that there is no conflict of interest

regarding the publication of this paper.

VIII. ACKNOWLEDGMENT

The research leading to these results has received funding

from the Spanish Ministry of Economy and Competitive-

ness under grant TEC2014-60491-R (Project 5GNORM). This

work also was supported by the Spanish National Science

Council and ERFD funds under Project TEC2014-60258-C2-

2-R.

REFERENCES

[1] Gandalf. Monitoring and self-tuning of RRM parameters in a multi-system network. [Online]. Available: http://www.celtic-initiative.org/

[2] SOCRATES. Self-Optimisation and self-ConfiguRATion in wirelEssnetworkS. [Online]. Available: http://www.fp7-socrates.org/

[3] SEMAFOUR. Self-Management for Unified Heterogeneous RadioAccess Networks. [Online]. Available: http://fp7-semafour.eu/

[4] SESAME. Small Cells Coordination for Multi-tenancy and Edgeservices. [Online]. Available: https://5g-ppp.eu/sesame/

[5] SELFNET. Framework for Self-Organised Network Management inVirtualised and Software Defined Networks. [Online]. Available:http://www.cognet.5g-ppp.eu/cognet-in-5gpp/

[6] COGNET. Cognitive Networks. [Online]. Available: https://5g-ppp.eu/[7] O. Aliu, A. Imran, M. Imran, B. Evans, “A Survey of Self Organiza-

tion in Future Cellular Networks,” IEEE Communications Surveys &

Tutorials, no. 99, pp. 1–26, 2012.[8] Sistelbanda. SN4G SON. [Online]. Available: http://sistelbanda.es/[9] Qualcomm. Self managing and enabling seamless roaming. [Online].

Available: https://www.qualcomm.com/videos/qualcomm-wi-fi-son[10] Huawei. Huawei’s Innovative Single SON Solution. [Online].

Available: http://www1.huawei.com/en/solutions/broader-smarter/[11] A. communications. Powering 4G networks. [Online]. Available:

http://www.airhopcomm.com

[12] Cellwize. Driving value through SON. [Online]. Available:http://www.cellwize.com

[13] AVIAT. Wireless products for small cell applications. [Online].Available: https://startupgenome.co/aviat-networks

[14] S. C. Forum. Small cell industry award. [Online]. Available:http://www.smallcellforum.org/events/awards

[15] 3GPP, “Technical Specification Group Services and System Aspects;Telecommunication Management; Self-Organizing Networks (SON);Concepts and requirements (Release 13),” 3GPP, TS 32.500, v13.0.0,2016.

[16] A. Imran, A. Zoha, A. Abu-Dayya, “Challenges in 5G: How toEmpower SON with Big Data for Enabling 5G,” IEEE Network, vol. 28,no. 6, pp. 27–33, 2014.

[17] NOKIA, “Intelligent Self-Organizing Networks: Increased automationfor higher network performance with lower costs,” White paper, 2013.

[18] N. Baldo, L. Giupponi, J. Mangues-Bafalluy, “Big Data EmpoweredSelf Organized Networks,” 20th IEEE European Wireless, 2014.

[19] B. Romanous, N. Bitar, A. Imran, H. Refai, “Network Densification:Challenges and Opportunities in enabling 5G,” 20th IEEE International

Workshop on Computer Aided Modelling and Design of Communica-tion Links and Networks (CAMAD, Special Session: Dense and Elastic

Big Data Empowered SON, pp. 129–134, 2015.

[20] D. Falko, “A Study of Self-Organization Mechanisms in Ad Hocand Sensor Networks,” Elseiver Computer Communications, pp. 3018–3019, 2008.

[21] D. Kreutz, F. M. V. Ramos, P. E. Verissimo, C. E. Rothenberg, S.Azodolmolky, S. Uhlig, “Software-Defined Networking: A Compre-hensive Survey,” Proceedings of the IEEE, vol. 103, no. 1, pp. 14–76,Jan 2015.

[22] 3GPP, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN); Self configuring and self optimizing network uses case andsolutions (Release 9),” 3GPP, TR T36.902, v9.2.0, 2010.

[23] ——, “Telecommunication management; Self-Organizing Networks(SON); Concepts and requirements (Release 11),” 3GPP, TS 32.500,v11.1.0, 2011.

[24] 3GPP, “Self Configuration of Network Elements; Concepts and Re-quirements, Release 9,” 3GPP, TS 32.501, 2010.

[25] 3GPP, “Self-configuration of network elements Integration ReferencePoint (IRP); Information Service (IS) (Release 10),” 3GPP, TS 32.502,v10.1.0, 2010.

[26] ——, “Self-Configuration of Network Elements Integration ReferencePoint (IRP); Common Object Request Broker Architecture (CORBA)Solution Set (SS)(Release 8),” 3GPP, TR 32.503, 2008.

[27] ——, “Software Management (SWM); Software and Integration Refer-ence Point (IRP) Requirements (Release 10),” 3GPP, TR 32.531, 2010.

[28] ——, “Software management Integration Reference Point (IRP); Infor-mation Service (IS) (Release 8),” 3GPP, TR 32.532, 2009.

[29] ——, “Software Management (SWM) Integration Reference Point(IRP): Common Object Request Broker Architecture (CORBA) So-lution Set (SS) (Release 9),” 3GPP, Tech. Rep. 32.533, 2009.

[30] ——, Tech. Rep.

[31] ——, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN) Network Resource Model (NRM) Integration Reference Point(IRP); Information Service (IS) (Release 11),” 3GPP, TS 32.762,v11.7.0, 2014.

[32] ——, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN) Network Resource Model (NRM) Integration Reference Point(IRP): CORBA solution set (Release 9),” 3GPP, TR 32.763, 2009.

[33] ——, “Evolved Universal Terrestrial Radio Access Netwrok (E-UTRAN) Network Resource Model (NRM) Integration Reference Point(IRP): eXtensible Markup Language (XML) definitions (Release 9),”3GPP, TR 32.765, v9.6.0, 2004.

[34] ——, “Study of Self-Organizing Networks (SON) related Operations,Administration and Maintenance (OAM) for Home Node B (HNB)(Release 9),” 3GPP, TR 32.821, v9.0.0, 2004.

[35] ——, “Telecommunication management; Self-Organizing Networks(SON); Study on self-healing (Release 9),” 3GPP, TR 32.823, 2010.

[36] 3GPP, “Performance measurements Evolved Universal Terrestrial Ra-dio Access Network (E-UTRAN)(Release 10),” 3GPP, TS 32.425,v10.6.0, 2012.

[37] 3GPP, “Universal Mobile Telecommunications System (UMTS);UTRAN Iu Interface RANAP Signalling,” 3GPP, TS 25.413, v3.4.0,1999.

[38] ——, “Radio Access (E-UTRA) and Evolved Universal TerrestrialRadio Access Network (E-UTRAN); Overall description; Stage 2(Release 12),” 3GPP, TS 36.300, 2015.

20

[39] ——, “LTE; Evolved Universal Terrestrial Radio Access Network (E-UTRAN); S1 Application Protocol (S1AP) (Release 9),” 3GPP, TS36.413, v9.4.0, 2010.

[40] 3GPP, “Technical Specification Group Radio Access Network; EvolvedUniversal Terrestrial Radio Access Network (EUTRAN); X2 Applica-tion Protocol (X2AP) (Release 11),” 3GPP, TS 36.423, v10.7.0, 2013.

[41] 3GPP, “Self-Organizing Networks (SON) Policy Network ResourceModel (NRM) Integration Reference Point (IRP); Information Ser-vice(IS), 3rd Generation Partnership Project (3GPP), (Release 10),”3GPP, TS 32.522, 2011.

[42] ——, “Self-Organizing Networks (SON): Policy Network ResourceModel (NRM) Integration Reference Point (IRP); Information Ser-vice(IS); Solution Set (SS) definitions, (Release 10),” 3GPP, TS 32.526,2011.

[43] ——, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN); Network Resource Model (NRM) Integration ReferencePoint (IRP); Solution Set(SS) definitions, (Release 10),” 3GPP, TS32.766, 2011.

[44] ——, “Telecommunication management; Self-Organizing Networks(SON); Self-healing concepts and requirements (Release 12),” 3GPP,TS 32.541, v12.0.0, 2014.

[45] ——, “Technical Specification Group Services and System Aspects;Telecommunication Management; Study on Energy Savings Manage-ment (ESM) (Release 10),” 3GPP, TR 32.826, 2010.

[46] ——, “Technical Specification Group Services and System Aspects;Study on Operations, Administration and Maintenance (OAM) aspectsof inter-Radio-Access-Technology (RAT) energy saving (Release 11),”3GPP, TR 32.834, 2012.

[47] ——, “Universal Mobile Telecommunications System (UMTS); LTE;Telecommunication management; Energy Saving Management (ESM);Concepts and requirements (Release 10),” 3GPP, TS 32.551, v10.1.0,2011.

[48] ——, “Technical Specification Group Radio Access Network; EvolvedUniversal Terrestrial Radio Access (E-UTRA); Radio Resource Control(RRC); Protocol Specification (Release 10),” 3GPP, TS 36.331, v10.7.0,2010.

[49] ——, “Performance Management (PM); Performance MeasurementsUTRAN (Release 9),” 3GPP, TS 32.405, 2009.

[50] ——, “Automatic Neighbor Relation (ANR) management, Conceptsand Requirements, (Release 11),” 3GPP, TR 32.511, 2011.

[51] 3GPP, “Technical Specification Group Services and System Aspects;Telecomunications Management; Self-Organizing Networks (SON)Policy Network Resource Model (NRM) Integration Reference Point(IRP)(Release 10),” 3GPP, TR 32.521, v10.2.0, 2009.

[52] 3GPP, “Configuration Management (CM); UTRAN network re-sources Integration Reference Point (IRP); Network Resource Model(NRM)(Release 9),” 3GPP, TS 322.642, 2009.

[53] ——, “TSG SA WG5 (Telecom Management) Meeting 85; Study ofimplementation alternative for SON coordination,” 3GPP, Tech. Rep.S5-122330, 2012.

[54] ——, “Configuration Management (CM); UTRAN network resourcesIntegration Reference Point (IRP); Solution Set (Release 10),” 3GPP,Tech. Rep. 32.646, v10.3.0, 2011.

[55] ——, “Radio Resource Control (RRC); Protocol specification (Release10),” 3GPP, TS 25.331, v10.6.0, 2011.

[56] ——, “UTRAN overall description (Release 10),” 3GPP, TS 25.401,2011.

[57] ——, “UTRAN Iu Interface: General Aspects and Principles (Release10),” 3GPP, TS 25.410, 2011.

[58] ——, “UTRAN Iur interface Radio Network Subsystem ApplicationPart (RNSAP) signalling (Release 9), institution = 3GPP,” TS 25.423,2010.

[59] ——, “Study on Network Management (NM) centralized Coverageand Capacity Optimization (CCO) Self-Organizing Networks (SON)function,” 3GPP, TS 32.836, 2013.

[60] ——, “Integration Reference Point (IRP) overview and usage guide(Release 10),” 3GPP, TS 32.103, v10.0.0, 2011.

[61] ——, “Self-Organizing Networks (SON) Policy Network ResourceModel (NRM) Integration Reference Point (IRP); Requirements (Re-lease 11),” 3GPP, TS 28.627, v11.1.0, 2012.

[62] ——, “Self-Organizing Networks (SON) Policy Network ResourceModel (NRM) Integration Reference Point (IRP); Information Service(IS) (Release 11),” 3GPP, TS 28.628, v11.4.0, 2014.

[63] ——, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN) Network Resource Model (NRM) Integration Reference Point(IRP); Information Service (IS) (Release 11),” 3GPP, TS 28.658,v11.0.0, 2012.

[64] ——, “Evolved Universal Terrestrial Radio Access Network (E-UTRAN) Network Resource Model (NRM) Integration Reference Point(IRP); Solution Set (SS) definitions (Release 11),” 3GPP, TS 28.659,2012.

[65] ——, “Gap analysis between 3GPP SA5 specifications and NGMN TopOperational Efficiency (OPE) Recommendations (Release 12),” 3GPP,TS 32.508, 2013.

[66] ——, “Data formats for multi-vendor plug and play eNodeB connectionto the network (Release 12),” 3GPP, TS 32.509, 2013.

[67] ——, “Gap analysis between 3GPP SA5 specifications and NGMN TopOperational Efficiency (OPE) Recommendations (Release 12),” 3GPP,TR 32.838, 2013.

[68] ——, “Study of heterogeneous networks management (Release 12),”3GPP, TR 32.835, v12.0.0, 2014.

[69] ——, “Study on Operations, Administration and Maintenance (OAM)aspects of Network Sharing (Release 12),” 3GPP, TR 32.851, v12.1.0,2013.

[70] ——, “Study on Next Generation Self-Optimizing Network (SON) forUTRAN and E-UTRAN (Release 12) ,” 3GPP, TR 37.822, 2013.

[71] ——, “Study on Energy Saving Enhancement for E-UTRAN (Release12),” 3GPP, TR 36.887, 2013.

[72] ——, “Study on Enhancements of OAM aspects of distributed Self-Organizing Networks (SON) functions, Release 12,” 3GPP, TR 32.860,2014.

[73] ——. Overview of 3gpp release 14.

[74] ——. (2016) RAN Evolution of LTE in Release 14. [Online]. Available:http://www.3gpp.org/news-events/3gpp-news/1768-ran rel14

[75] ——, “Technical Specification Group Services and System Aspects;Telecommunication Management; Project scheduling and open issuesfor SA5, (Release 8),” 3GPP, Tech. Rep. 30.818, v8.0.8, 2010.

[76] M. Amirijoo, P. Frenger, F. Gunnarsson, “Neighbor cell relation list andphysical cell identity self-organization in LTE,” ICC Workshops-IEEE

International Conference on Communications Workshops, pp. 31–41,March 2008.

[77] F, Parodi, M, Kylvaja, G, Alford, J, Li and J, Pradas, “An AutomaticProcedure for Neighbor Cell List Definition in Cellular Networks,”IEEE International Symposium on a World of Wireless, Mobile andMultimedia Networks (WoWMoM), pp. 1–6, June 2007.

[78] 3GPP, “Technical Specification Group Radio Access Network, EvolvedUniversal Terrestrial Radio Access Network (E-UTRAN), Self config-uring and Self optimizing Network (SON) Uses Case and Solutions(Release 9),” 3GPP, TR 36.902, v9.3.1, 2011.

[79] F. Parodi, M. Kylvaja, G. Alford, J. Li, J. Pradas, “RecommendedPractices for Multi-vendor SON Deployment,” D2- NGMN AlianceV1.0, January 2014.

[80] 3GPP, “Evolved Universal Terrestrial Radio Access (E-UTRA); Phys-ical layer procedures,” 3GPP, TS 36.213, v10.2.0, 2011.

[81] NGMN, “Recommended Practices for Multi-vendor SON Deploy-ment,” NGMN P-SmallCell Work Stream 2 (WS2) first deliverable,2014.

[82] 3GPP, “Evolved Universal Terrestrial Radio Access (E-UTRA) andEvolved Universal Terrestrial Radio AccessE-UTRAN; Overall de-scription; Stage 2,” 3GPP, TS 36.300 v12.6.0, 2015.

[83] ——, “SON self-healing management,” 3GPP, Tech. Rep. S5-460036,2011.

[84] Seppo Hamalainen, Henning Sanneck, Cinzia Sartori, LTE Self-

Organising Networks (SON): Network Management Automation for

Operational Efficiency. John Wiley and Sons, 2012.

[85] L. C. Schmelz, M. Amirijoo, A. Eisenblaetter, R. Litjens, M. Neuland,J. Turk, “A Coordination Framework for Self-Organisation in LTENetworks,” IEEE International Symposium on Integrated NetworkManagement, 2011.

[86] Z. Altman, M.Amirijoo, F.Gunnarson, H.Hoffmann, I.Z. Kovacs, D.Laselva, B. Sas, K. Spaey, A. Tall, H. Van Den Berg, K. Zetterberg,“On design principles for self-organizing network functions,” 11th In-ternational Symposium on Wireless Communications Systems (ISWCS),pp. 454–459, 2014.

[87] 3GPP, “Study on Minimization of drive-tests in Next GenerationNetworks,” 3GPP, TR 36.805, 2009.

[88] J. Johansson, W. A. Hapsari, S. Kelley, G. Bodog, “Minimization ofdrive tests in 3GPP (Release 11),” IEEE Communications Magazine,November 2012.

[89] Nokia Siemens Networks. (2012) Automating Core NetworkManagement, Mobile World Congress. [Online]. Available:https://networks.nokia.com/solutions/cloud-native-core-network

21

[90] A. Blenk, A. Basta, M. Reisslein, W. Kellerer, “Survey on NetworkVirtualization Hypervisors for Software Defined Networking,” IEEECommunications Surveys Tutorials, vol. 18, no. 1, pp. 655–685, 2016.

[91] R. Caruana, A. Niculescu-Mizil, “An empirical comparison of super-vised learning algorithms,” 23rd International Conference on Machine

Learning, 2006.[92] P. McCullagh, J. Nelder, “Generalized Linear Models, Second Edition,”

Boca Raton: Chapman and Hall, 1989.[93] H. Zhang, “The Optimality of Naive Bayes,” 17th International Florida

Artificial Intelligence Research Society Conference, 2004.[94] V. Vapnik, “An Overview of Statistical Learning Theory,” IEEE Trans-

actions on Neural Networks, vol. 10, no. 5, 1999.[95] A. J. Smola, B. Scholkopf, “A tutorial on support vector regression,”

Statistics and Computing, pp. 199–222, 2004.[96] A. Muller, “Kernel Approximations for Efficient SVMs (and other

feature extraction methods),” Data Science & Machine Learning Re-

sources, 2012.[97] Bishop M.C., “Pattern recognition and machine learning,” Busines Dia,

Llc. Springer Science, 2006.[98] J. R. Quinlan, “Induction of Decision Trees. Machine Learning,”

Kluwer Academic Publishers, pp. 81–106, 1986.[99] L. Rokach, O. Maimon, “Data Mining with Decision Trees: Theory and

Applications,” Series in Machine Perception and Artificial Intelligence,2008.

[100] D. Opitz, R. Maclin, “Popular Ensemble Methods: An EmpiricalStudy,” Journal of Artificial Intelligence Research 11, pp. 169–198,1999.

[101] L. Rokach, “Ensemble-based classifiers,” Artificial Intelligence, pp. 1–39, 2010.

[102] T. G. Dietterich, “An Experimental Comparison of Three Methodsfor Constructing Ensembles of Decision Trees: Bagging, Boosting andRandomization,” Machine Learning, vol. 40, pp. 139–157, 2000.

[103] ——, “Machine-Learning Research: Four Current Directions,” Ameri-

can Association for Artificial Intelligence, 1997.[104] Y. Freund, R. E. Schapire, “A Decision-Theoretic Generalization of On-

Line Learning and Application to Boosting,” 2nd European Conference

on Machine Learning, pp. 148–156, 1995.[105] G. Zoubin, “Unsupervised Learning,” Gatsby Computational Neuro-

science Unit University College London, UK, 2004.[106] W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery, “Section

16.1. Gaussian Mixture Models and k-Means Clustering,” The Art ofScientific Computing (3rd ed.); York: Cambridge University, 2007.

[107] A. Ultsch, H. P. Siemon, “Kohonen’s Self Organizing Feature Maps forExploratory Data Analysis,” International Neural Network Conference,1990.

[108] T. Hastie, R. Tibshirani, J. Friedman, The Elements of Statistical

Learning (2nd ed.). New York: Springer, 2009.[109] R. Nock, F. Nielsen, “On Weighting Clustering,” IEEE Trans on Pattern

Analysis and Machine Intelligence, vol. 28, pp. 1–13, 2006.[110] J. Moysen, N. Baldo, L. Giupponi, J. Mangues-Bafalluy, “Predicting

QoS in LTE HetNets based on location-independent UE measurement,”20th IEEE International Workshop on Computer Aided Modelling andDesign of Communication Links and Networks, Guildford, UK, 2015.

[111] J. Moysen, L. Giupponi, J. Mangues-Bafalluy, “A Mobile NetworkPlanning tool based on Data Analytics,” Mobile Information Systems,

Hindawi, 2017.[112] S. T. Roweis, K. S. Lawrence, “Nonlinear Dimensionality Reduction

by Locally Linear Embedding,” American Association for the Advance-

ment of Science, pp. 2323–2326, 2000.[113] S. T. Roweis, “EM algorithms for PCA and SPCA,” Advances in Neural

Information Processing Systems, The MIT Press, 1998.[114] V. J. Hodge, J. Austin, “A Survey of Outlier Detection Methodologies,”

Artificial Intelligence Review, vol. 22, 2004.[115] I. de la Bandera, R. Barco, P. Munoz, I. Serrano, “Cell Outage De-

tection Based on Handover Statistics,” IEEE Communications Letters,vol. 19 (7), 2015.

[116] P. Munoz, R. Barco, I. Serrano, A. Gomez Andrades, “CorrelationBased Time-Series Analysis for Cell Degradation,” IEEE Communica-

tions Letters, 2016.[117] S. Schaal and C. Atkeson, “Robot juggling: An implementation of

memory-based learning,” Control Systems Magazine, 1994.[118] S. Thrun, “Learning to play the game of chess,” Advances in Neural

Information Processing Systems 7, Cambridge, MA, 1995. The MIT

Press., 1995.[119] M. L. Littman, “Markov games as a framework for multi-agent

reinforcement learning,” 11th International Conference on Machine

Learning, pp. 157–163, 1994.

[120] P. Liviu and L. Sean, “Cooperative Multi-Agent Learning: The State ofthe Art,” Autonomous Agents and Multi-Agent Systems, vol. 11, no. 3,pp. 387–434, 2005.

[121] R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction.MIT Press, 1998.

[122] 3GPP, “Technical Specification Group Services and Systems Aspects;Charging management; Charging Data Record (CDR) parameter de-scription (Release 13),” 3GPP, TS 32.298, v13.3.0, 2016.

[123] ——, “Telecommunication management; performance management(pm); concept and requirements (release 10),” 3GPP, TS 32.401,v10.2.0, 2010.

[124] Internet Engineering Task Force (IETF), “Information Model for IPFlow Information Export (IPFIX),” 3GPP, Tech. Rep. RFC 7012, 2013.

[125] 3GPP, “Technical Specification Group TSG RAN; Radio measurementcollection for Minimization of Drive Tests (MDT), Release 10,” 3GPP,TS 37.320, v2.0.0, 2010.

[126] Orange. The D4D Challenge. [Online]. Available:http://www.d4d.orange.com/en/Accueil

[127] 3GPP, “Technical Specification Group Core Network and Terminals;AT command set for User Equipment (UE) (Release 12),” 3GPP, TS27.007 v12.1.0, 2013.

[128] OpenCellID. Largest Open Database of Cell Towers and Geolocation.[Online]. Available: https://opencellid.org/

[129] OpenSignal. 3G and 4G LTE Cell Coverage Map. [Online]. Available:https://opensignal.com/

[130] Antenas GSM. [Online]. Available: http://antenasgsm.com/

[131] Google geolocation api. [Online]. Available:https://developers.google.com/maps/

[132] N. Bui, J. Widmer, “Owl: a reliable online watcher for LTE controlchannel measurements,” arXiv preprint arXiv:1606.00202l, 2016.

[133] H. D. Trinh, N. Bui, J. Widmer, L. Giupponi, P. Dini, “Analysis andModeling of Mobile Traffic Using Real Traces,” submitted to IEEE

Personal Indoor and Mobile Radio Communications (PIMRC), 2017.

[134] M. Peng, D. Liang, Y. Wei, J. Li, and H. H. Chen, “Self-configurationand self-optimization in LTE-advanced heterogeneous networks,” IEEE

Communications Magazine, vol. 51, no. 5, pp. 36–45, 2013.

[135] S. S. Mwanje, A. Mitschele-Thiel, “Minimizing Handover PerformanceDegradation Due to LTE Self Organized Mobility Load Balancing,”77th IEEE Conference on Transactions on Vehicular Technology (VTC

Spring), pp. 1–5, 2013.

[136] P. Munoz, R. Barco, I. de la Bandera, “Optimization of load balancingusing fuzzy Q-Learning for next generation wireless networks,” Expert

Systems with Applications, Elsevier, vol. 40, pp. 984–994, 2013.

[137] P. Munoz, R. Barto, I. de la Bandera, “Load Balancing and Handoverjoint optimization in LTE networks using Fuzzy logic ReinforcementLearning,” Computer Networks Elsevier, 2014.

[138] J. Suga, Y. Kojima, M. Okuda, “Centralized Mobility Load BalancingScheme in LTE Systems,” 8th International Symposium on Wireless

Communication Systems, Aachen, 2011.

[139] E. Bergner, “Unsupervised Learning of Traffic Patterns in Self-Optimizing 4th Generation Mobile Networks,” Master of Science The-sis, KTH Computer Science and Communications, Stockholm, Sweden,2012.

[140] C. A. S. Franco, J. R. B. de Marca, “Load balancing in self-organizedheterogeneous LTE networks: A statistical learning approach,” IEEE

Latin-American Conference on Communications (LATINCOM), pp. 1–5, 2015.

[141] W. Qin, Y. Teng, M. Song, Y. Zhang, and X. Wang, “A Q-learning ap-proach for mobility robustness optimization in LTE-SON,” 15th IEEE

International Conference on, Communication Technology (ICCT), pp.818–822, 2013.

[142] S. S. Mwanje, A. Mitschele-Thiel, “Distributed cooperative Q-learningfor mobility-sensitive handover optimization in LTE SON,” IEEE

Symposium the Computers and Communication (ISCC), pp. 1–6, 2014.

[143] P. Munoz, R. Barco, and I. de la Bandera, “On the Potential of Han-dover Parameter Optimization for Self-Organizing Networks,” IEEE

Transactions on Vehicular Technology, 2013.

[144] N. Sinclair, D. Harle, I. A. Glover, J. Irvine, R. C. Atkinson, “ParameterOptimization for LTE Handover using an Advanced SOM Algorithm,”IEEE Conference on Transactions on Vehicular Technology (VTC),2013.

[145] N. Sinclair, D. Harle, I. Glover, J. Irvine, and R. Atkinson, “Anadvanced SOM algorithm applied to handover management withinLTE,” IEEE Transactions on Vehicular Technology, pp. 183–1894,2013.

22

[146] H. Farooq, A. Imran, “Spatio-temporal Mobility Prediction in ProactiveSelf-Organizing Cellular Networks,” IEEE Communications Letters,vol. 21, no. 2, pp. 370–373, 2017.

[147] Z. Ali, N. Baldo, J. Mangues-Bafalluy, and L. Giupponi, “Machinelearning based handover management for improved qoe in LTE,”IEEE/IFIP Network Operations and Management Symposium, pp. 794–798, 2016.

[148] E. Ostlin, H. J. Zepernick, H. Suzuki, “Macrocell radio wave propa-gation prediction using an artificial neural network,” IEEE Vehicular

Technology Conference (VTC Fall), p. 57, 2004.

[149] A. Quintero and O. Garcıa, “A profile-based strategy for managing usermobility in third-generation mobile systems,” IEEE CommunicationsMagazine, pp. 134–139, 2004.

[150] K. Majumdar and N. Das, “Mobile user tracking using a hybrid neuralnetwork,” in Wireless Networks, pp. 275–284, 2005.

[151] R. Razavi, S. Klein, H. Claussen, “A fuzzy reinforcement learningapproach for self-optimization of coverage in LTE networks,” Bell Labs

Tachnical Journal, vol. 15, pp. 153–175, 2010.

[152] M. Naseer ul Islam, A. Mitschele-Thiel, “A Cooperative Fuzzy Q-learning for Self-Organized Coverage and Capacity Optimization,”IEEE 23rd International Symposium on Personal Indoor and MobileRadio Communications (PIMRC), Sept 2012.

[153] J. Li, J. Zeng, X. Su, W. Luo, J. Wang, “Self-Optimization of Cov-erage and Capacity in LTE Networks Based on Central Control andDecentralized Fuzzy Q-Learning,” International Journal of Distributed

Sensor Networks, Hindawi, 2012.

[154] S. Fan, H. Tian, and C. Sengul, “Self-optimization of coverage andcapacity based on a fuzzy neural network with cooperative reinforce-ment learning,” Journal on Wireless Communications and Networking,2014.

[155] A. Galindo-Serrano, L. Giupponi, G. Auer, “Distributed Learningin Multiuser OFDMA Femtocell Networks,” 73rd IEEE Vehicular

Technology Conference (VTC Spring), pp. 1–6, 2011.

[156] M. Dirani and Z. Altman, “A cooperative reinforcement learningapproach for inter-cell interference coordination in OFDMA cellularnetworks,” 8th Int Modeling and Optimization in Mobile, Ad Hoc andWireless Networks (WiOpt), pp. 170–176, 2010.

[157] M. Bennis, S. M. Perlaza, P. Blasco, “Self-Organization in Small CellNetworks: A Reinforcement Learning Approach,” IEEE Transactions

on Wireless Communications, vol. 12, no. 7, 2013.

[158] M. Simsek, A. Czylwik, A. Galindo-Serrano, L. Giupponi, “Improveddecentralized Q-learning algorithm for interference reduction in LTE-femtocells,” 5th International Workshop on Self-Organizing Networks

IWSON, Glasgow, 2015.

[159] M. Miozzo, L. Giupponi, M. Rossi, P. Dini, “Distributed Q-Learningfor Energy Harvesting Heterogeneous Networks,” IEEE ICC 2015Workshop on Green Communications and Networks with Energy Har-

vesting, Smart Grids, and Renewable Energies, June 2015.

[160] A. Dudnikova, P. Dini, L. Giupponi, D. Panno, “Multiple Criteria plusFuzzy Logic Switch Off Method for Dense Heterogeneous Networks,”20th IEEE International Workshop on Computer Aided Modelling and

Design of Communication Links and Networks, Guildford (UK), 2015.

[161] E. Ternon, P. Agyapong, L. Hu, A. Dekorsy, “Energy Savings in Het-erogeneous Networks with Clustered Small Cell Deployments,” 11thIEEE International Symposium on Wireless Communications Systems

(ISWCS), 2014.

[162] E. Ternon, P. Agyapong, L. Hu, and A. Dekorsy, “Database-aidedEnergy Savings in Next Generation Dual Connectivity HeterogeneousNetworks,” IEEE WCNC’14 Track 3: Mobile and Wireless Networks,2014.

[163] J. Moysen and L. Giupponi, “A Reinforcement Learning based solutionfor Self-Healing in LTE networks,” 80th IEEE Vehicular Technology

Conference (VTC Fall), Vancouver, Canada, 2014.

[164] O. Onireti, A. Zoha, J. Moysen, A. Imran, L. Giupponi, M. AliImran, A. Abu-Dayya, “A Cell Outage Management Framework forDense Heterogeneous Networks,” IEEE Transactions on Vehicular

Technology, vol. 64, pp. 2097–2113, April 2015.

[165] A. Saeed, O. G. Aliu, M. A. Imran, “Controlling Self Healing CellularNetworks using Fuzzy Logic,” IEE Wireless Communications andNetworking Conference (WCNC), Paris, France, April 2012.

[166] F. Chernogorov, J. Turkka, T. Ristaniemi, A. Averbuch, “Detection ofSleeping Cells in LTE Networks Using Diffusion Maps,” IEEE 73rd

Vehicular Technology Conference (VTC Spring), pp. 1–5, may 2011.

[167] E. J. Khatib, R. Barco, A. Gomez-Andrades, I. Serrano, “Diagnosisbased on genetic fuzzy algorithms for LTE Self-Healing,” IEEE Trans-

actions on Vehicular Technology, 2015.

[168] R. M. Khanafer, B. Solana, J. Triola, R. Barco, L. Moltsen, Z. Altman,P. Lazaro, “Automated Diagnosis for UMTS Networks Using BayesianNetwork Approach,” IEEE Transactions on Vehicular Technology,vol. 57, 2008.

[169] G. F. Ciocarlie, S. N. Sanneckaczki, “Detecting Anomalies in CellularNetworks Using an Ensemble Method,” 9th CNSM and Workshops,2013.

[170] A. Multazamah, S. Navrati, R. Abhishek, “Efficient Cell OutageDetection in 5G Het-Nets Using Hidden Markov Model,” IEEE Com-

munications Letters, vol. 20, no. 3, 2016.

[171] W. Xue, M. Peng, Y. Ma, and H. Zhang, “Classification-based approachfor cell outage detection in self-healing heterogeneous networks,” IEEE

Wireless Communications and Networking Conference (WCNC), p.28222826, 2014.

[172] A. Zoha, A. Saeed, A. Imran, M. A. Imran, and A. Abu-Dayya, “A SONsolution for sleeping cell detection using low-dimensional embeddingof mdt measurements,” IEEE 25th Annual International Symposium

on Personal, Indoor, and Mobile Radio Communication (PIMRC), pp.1626–1630, 2014.

[173] S. Chernov, F. Chernogorov, D. Petrov, T. Ristaniemi, “Data miningframework for random access failure detection in LTE networks,” 25th

IEEE Annual International Symposium on Personal, Indoor, and Mobile

Radio Communication (PIMRC), pp. 1321–1326, 2014.

[174] R. Barco, V. Wille, and L. Dıez, “System for automated diagnosisin cellular networks based on performance indicators,” EuropeanTransactions on Telecommunications, pp. 399–409, 2005.

[175] R. M. Khanafer, B. Solana, J. Triola, R. Barco, L. Moltsen, Z.Altman, P. Lazaro, “Automated diagnosis for UMTS networks usingBayesian network approach,” IEEE Transactions on Vehicular Technol-

ogy, vol. 57, p. 24512461, 2008.

[176] H. Y. Lateef, A. Imran, A. Abu-dayya, “A Framework for Classificationof Self-Organising Network Conflicts and Coordination Algorithms,”IEEE Personal Indoor and Mobile Radio Communications PIMRC,2013.

[177] O. Iacoboaeia, B. Sayrac, S. Jemaa, P. Bianchi, “SON Coordination forparameter conflict resolution: A reinforcement learning framework,”IEEE Wireless Communications and Networking Conference 2014(IEEE WCNC 2014), April 2014.

[178] P. Munoz, R. Barco, and S. Flores, “Conflict Resolution BetweenLoad Balancing and Handover Optimization in LTE Networks,” IEEE

Communications Letters, vol. 18, pp. 1795–1798, 2014.

[179] J. Moysen and L. Giupponi, “Self-coordination of parameter con-flicts in D-SON architectures: a Markov decision process framework,”EURASIP Journal on Wireless Communications and Networking, 2015.

[180] F. Chernogorov and T. Nihtila, “QoS Verification for Minimizationof Drive Tests in LTE Networks,” 75th IEEE Vehicular Technology

Conference (VTC Spring), pp. 6–9, may 2012.

[181] F. Chernogorov and J. Puttonen, “User satisfaction classification forMinimization of Drive Tests QoS verification,” 24th IEEE Personal

Indoor and Mobile Radio Communications (PIMRC), London, United

Kingdom, pp. 2165–2169, September 2013.

[182] J. Moysen, L. Giupponi, J. Mangues-Bafalluy, “On the Potential of En-semble Regression Techniques for Future Mobile Network Planning,”21th IEEE Symposium on Computers and Communications (ISCC),

Messina, Italy, 2016.

[183] ——, “A Machine Learning enabled network Planning tool,” 27thIEEE Personal Indoor and Mobile Radio Communications (PIMRC),

Valencia, Spain, 2016.

[184] B. Daroczy, A. Benczur, P. Vaderna, “Machine Learning based sessiondrop prediction in LTE networks and its SON aspects,” 5th Inter-

national Workshop on Self-Organizing Networks (IWSON), Glasgow,2015.

[185] J. Turkka, D. Gil, “Anomaly Detection Framework for Tracing Prob-lems in Radio Networks,” 10th International Conference on Networks

(ICN), 2011.

[186] R. Romeikat, B. Bauer, T. Bandh, G. Carle, H. Sanneck, L. C.Schmelz, “Policy-driven workflows for mobile network managementautomation,” 6th International Wireless Communications and Mobile

Computing Conference (IWCMC), pp. 1111–1115, 2010.

[187] V. Buenestado, M. Toril, S. Luna, J. M. Ruiz, A. Mendo, “Self-tuningof remote electrical tilts based on call traces for coverage and capacityoptimization in LTE,” IEEE Transactions on Vehicular Technology,2017.

[188] M. Toril, R. Acedo-Hernandez, S. Luna-Ramırez, A. Sanchez, C.Ubeda, “Automatic clustering algorithms for indoor site selection inLTE,” Hindawi Mobile Information Systems, 2017.

23

[189] A. Galindo-Serrano, B. Sayrac, S. Ben Jemaa, J. Riihijarvi and P.Mahonen, “Harvesting MDT Data: Radio Environment Maps forCoverage Analysis in Cellular Networks,” International Conference on

Cognitive Radio Oriented Wireless Networks (CROWNCOM), 2013.[190] D. Clark, et all., “A Knowledge plane for the Internet,” in proc. of

the 2003 Conference on Applications, Technologies, Architectures, andProtocols for Computer Communications, 2003.

[191] L. M. Rios, N. V. Sahinidis, “Derivative free optimization: A reviewof algorithms and comparison of software implementations,” Journalof Global Optimization, pp. 1247–1293, 2013.

[192] H. Derbel, N. Agoulmine, M. Salan, “Anema: Autonomic networkarchitecture to support self-configuration and self-optimization in ipnetworks,” Computer Networks, pp. 418–430, 2009.

[193] e. a. M. Zorzi, “COBANETS: A new paradigm for cognitive communi-cations systems,” International Conference on Computing, Networking

and Communications ICNC, 2016.[194] A. Mestres, et all., “Knowledge Defined Networking,” arXiv preprint

arXiv:1606.06222.v2, 2016.[195] I. Ben Yahia, et all., “CogNitive 5G networks: Comprehensive operator

use cases with machine learning for management operations,” 20thConference on Innovations in Clouds, Internet and Networks (ICIN),2017.

[196] J. Bendriss, et all., “AI for SLA Management in ProgrammableNetworks,” International Conference on Design of Reliable Communi-cation Networks, 2017.

[197] F. Ricciato, P. Widhalm, M. Craglia and F. Pantisano, “Estimatingpopulation density distribution from network-based mobile phone data,

journal=JRC and AIT,” 2015.[198] M. Laner, P. Svoboda, P. Romirer, N. Nikaein, F. Ricciato, “A Compar-

ison Between One-Way Delays in Operating HSPA and LTE Networks,journal =8th Int’l workshop on Wireless Network Measurements (WIN-MEE’12), Pedeborn, Germany,” 2012.

[199] Centre Tecnologic de Telecomunicacions de Catalunya (CTTC).The LENA-EPC Network simulator. [Online]. Available:http://networks.cttc.es/mobile-networks/software-tools/lena/

[200] European Union Agency for Network and Information Security (enisa),“Privacy by design in big data,” An overview of privacy enhancing

technologies in the era of big data analytics, 2015.[201] K. A. Sallehab and L. Janczewski, “Technological, organizational and

environmental security and privacy issues of big data: A literaturereview,” Procedia Computer Science, ELSEVIER, pp. 19–28, 2016.

[202] Y. Bengio, A. Courville, P. Vincent, “Representation Learning: AReview and New Perspectives,” IEEE Transactions on Pattern Analysis

and Machine Intelligence, vol. 35, no. 8, pp. 1798–1828, 2013.[203] I. Goodfellow, Y. Bengio, A. Courville, Deep Learning. MIT Press,

2016, http://www.deeplearningbook.org.[204] J. Hu and M. P. Wellman, “Nash Q-Learning for General-Sum Stochas-

tic Games,” Journal of Machine Learning Research, vol. 4, pp. 1039–1069, 2003.

[205] M. Zorzi, A. Zanella, A. Testolin, M. Filipo de Grazia, M. Zorzi,“Cognition-Based Networks: A New Perspective on Network Optimiza-tion Using Learning and Distributed Intelligence,” IEEE Access The

journal for rapid open access publishing, Special Section on Artificial

Intelligence Enabled Networking, vol. 3, pp. 1512–1530, 2015.